ℹ️ Skipped - page is already crawled
| Filter | Status | Condition | Details |
|---|---|---|---|
| HTTP status | PASS | download_http_code = 200 | HTTP 200 |
| Age cutoff | PASS | download_stamp > now() - 6 MONTH | 0.5 months ago |
| History drop | PASS | isNull(history_drop_reason) | No drop reason |
| Spam/ban | PASS | fh_dont_index != 1 AND ml_spam_score = 0 | ml_spam_score=0 |
| Canonical | PASS | meta_canonical IS NULL OR = '' OR = src_unparsed | Not set |
| Property | Value |
|---|---|
| URL | https://www.mdpi.com/1999-4907/17/4/419 |
| Last Crawled | 2026-03-28 11:24:23 (14 days ago) |
| First Indexed | not set |
| HTTP Status Code | 200 |
| Meta Title | Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022) |
| Meta Description | Hurricanes routinely damage forests across the North Atlantic Basin, yet efforts to characterize their impacts have had mixed subregional success. To elucidate these challenges, this study analyzed pre- and post-hurricane land surface phenology (LSP) for 44 moderate and strong hurricanes over 22 years using the Enhanced Vegetation Index (EVI). We statistically grouped storms based on their long-term climate attributes, then compared subregional impacts with wind speed and land cover. After accounting for wind speed, responses differed among the six subregions. The Southeast U.S. showed declines in EVI for the first winter and first year post storm, but this response was weak or absent elsewhere. The Central America region declined in the first winter but not in the subsequent growing season, while four other regions showed no increased impact with wind speed in either season. We then examined six category 4 hurricanes using a forest mask. In dry areas, drought-sensitive vegetation explained weak responses, whereas in the humid tropics, rapid refoliation or sprouting was common. These factors complicate optical remote sensing assessments. Rapid evaluations can mistake defoliation for more substantial damage, and delayed assessments can confuse EVI recovery with structural recovery. Results underscore the need for ecologically tailored monitoring approaches. |
| Meta Canonical | null |
| Boilerpipe Text | Search
for Articles
:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
Section
Special Issue
Volume
Issue
Number
Page
Logical Operator
Operator
Search Text
Search Type
add_circle_outline
remove_circle_outline
Article Menu
Font Type:
Arial
Georgia
Verdana
Font Size:
Aa
Aa
Aa
Line Spacing:
Column Width:
Background:
Open Access
Feature Paper
Article
by
Carlos Topete-Pozas
and
Steven P. Norman
*
US Department of Agriculture Forest Service, Southern Research Station, Eastern Forest Environmental Threat Assessment Center, 200 WT Weaver Blvd, Asheville, NC 28804, USA
*
Author to whom correspondence should be addressed.
Submission received: 24 February 2026
/
Revised: 17 March 2026
/
Accepted: 25 March 2026
/
Published: 27 March 2026
Abstract
Hurricanes routinely damage forests across the North Atlantic Basin, yet efforts to characterize their impacts have had mixed subregional success. To elucidate these challenges, this study analyzed pre- and post-hurricane land surface phenology (LSP) for 44 moderate and strong hurricanes over 22 years using the Enhanced Vegetation Index (EVI). We statistically grouped storms based on their long-term climate attributes, then compared subregional impacts with wind speed and land cover. After accounting for wind speed, responses differed among the six subregions. The Southeast U.S. showed declines in EVI for the first winter and first year post storm, but this response was weak or absent elsewhere. The Central America region declined in the first winter but not in the subsequent growing season, while four other regions showed no increased impact with wind speed in either season. We then examined six category 4 hurricanes using a forest mask. In dry areas, drought-sensitive vegetation explained weak responses, whereas in the humid tropics, rapid refoliation or sprouting was common. These factors complicate optical remote sensing assessments. Rapid evaluations can mistake defoliation for more substantial damage, and delayed assessments can confuse EVI recovery with structural recovery. Results underscore the need for ecologically tailored monitoring approaches.
1. Introduction
The North Atlantic Basin (NAB) commonly experiences extreme hurricanes with severe consequences for the natural and human systems of the region [
1
,
2
,
3
,
4
]. Notably, the intensity and frequency of hurricanes in this region have been changing over the recent decades [
3
,
5
,
6
]. Between 1989 and 2018, an estimated one million km
2
of forest and other land cover types were affected by hurricanes in the NAB, representing one of the highest damaged areas globally [
7
]. Indeed, the majority of studies regarding hurricane impacts on forest (67%) were conducted in this region [
7
].
Remote sensing has become a critical tool for assessing hurricane impacts and subsequent vegetation recovery at different spatio-temporal scales [
8
,
9
]. Its usefulness for monitoring within the NAB is influenced by complex patterns of land cover and seasonal vegetation dynamics, both of which are informed by land surface phenology (LSP). Expanded use of LSP has improved how we understand disturbance impacts in light of seasonal and multi-year environmental and anthropogenic influences [
10
,
11
,
12
]. LSP-derived measures relate to important ecosystem processes, such as carbon and water cycles, timber markets, the fuels available for wildfire, and species interactions and connectivity [
13
]. LSP is particularly useful for tracking vegetation productivity, and this varies regionally and locally with vegetational composition, climate, soil, management practices, and weather disturbances, among other factors [
14
].
Recent research demonstrates the broad monitoring capabilities of LSP to characterize damage from hurricanes. Continuous satellite measurements give researchers the ability to track commonly used LSP parameters that capture key stages of seasonal vegetation behavior, such as the start of spring or fall or the peak of the growing season. In tropical or subtropical climates, hurricanes can impact the onset or progression of the late growing season contingent on when storms occur. This limitation makes LSP parameters that capture seasonal timing less reliable. Instead, for consistency, post-storm season integral phenometrics are more suitable, such as area under the curve (AUC) for all or a portion of the year. Such measures are recognized LSP phenometrics that provide a robust mechanism for understanding seasonal and annual variability [
10
,
15
,
16
,
17
]. Along the US coast, studies have demonstrated the value of short- and long-term monitoring using different sensors and spectral indices including the 10 m resolution Sentinel 2 Normalized Difference Vegetation Index (NDVI) [
18
], 30 m resolution Landsat 5 and 7 and NDVI [
19
], and five vegetation indices from MODIS 1 km 16-day [
20
]. Others [
21
] assessed the effects of Hurricane Maria in Puerto Rico using Landsat 8 data with spectral mixture and statistical analyses in association with landform characteristics and forest structure.
Enhanced Vegetation Index (EVI) has been considered by several authors as a more suitable index to monitor forest disturbance in tropical regions because of EVI’s sensitivity to canopy variation in high biomass regions where NDVI saturates [
22
,
23
,
24
]. Research has demonstrated the effectiveness of MODIS EVI to assess the extent of immediate impacts from Hurricane Dean in the tropical forest of the Yucatan Peninsula, Mexico [
25
], and in Central America after Hurricane Felix [
26
]. Others [
27
] used MODIS EVI to assess hurricane damage and recovery in the northern Gulf and found that impacts vary by event, year, vegetation characteristics, and climate attributes [
28
], and monitored forest recovery along the northern Gulf comparing MODIS EVI with other spectral indices. In the Caribbean, ref. [
29
] used a MODIS disturbance index to capture both drought and hurricane impacts over four large Caribbean islands, and stressed the importance of considering land cover. In Cuba, ref. [
30
] assessed the extent and severity of damage to mangroves after Hurricane Irma with integrated use of MODIS and Sentinel-2 data. MODIS has been useful for addressing hurricane impacts in the tropics because its high temporal frequency helps overcome issues with cloud cover while isolating immediate responses and longer-term implications.
The NAB’s complex meteorological and land cover characteristics present challenges for monitoring hurricane impacts as methods may not be uniformly effective [
31
,
32
]. While rapid optical remote sensing assessments are broadly effective for recognizing the footprint of potential impacts, commonly used techniques struggle with ecological precision, such as lasting structural damage and successional impacts in the tropics. Assessments in the humid tropics are especially time-sensitive with a high risk of misinterpreting recovery of the remote sensing index as something more. The most intensive research has taken place in the southern U.S., and this regional bias limits our understanding of monitoring effectiveness across the broader region [
7
]. The broader NAB region includes areas with substantially different vegetation and climate variability that may require different approaches for hurricane impact assessments to be accurate. Our research addresses these regional differences in the effectiveness of LSP monitoring after hurricanes to clarify important limitations and research needs. Our specific objectives include evaluating the importance of wind speed, land cover, and climate through the use of regional assessment and case studies.
2. Materials and Methods
2.1. Study Area and Hurricanes
We focused on major storms rather than tropical storms and Saffir–Simpson category 1 hurricanes to increase the likelihood of capturing wind damage to forests. We selected 44 category 2–5 hurricanes that made landfall within the NAB between the years 2001 to 2022 across the Southeast U.S., Mexico, the Caribbean, and Central America. We determined the landfall site and inland path for each hurricane using the International Best Track Archive for Climate Stewardship (IBTrACS) dataset [
33
]. For each hurricane, we defined a semi-uniform focal area by drawing a polygon that extended 25 km on each side of the track and 100 km inland. This standardized footprint is intended to emphasize the near-track region most likely to experience inner-core winds and associated severe wind damages and provides a consistent spatial basis for comparable impact estimation across storms and regions. In some cases, such as in the Caribbean, the polygons were somewhat smaller due to the size of the island or the need to exclude water bodies. We used each footprint to derive MODIS EVI data, land cover, and climate attributes.
Figure 1
shows an overview of the data used and workflow.
2.2. Climate Regionalization
Various researchers have classified the NAB-related tropical cyclones in subregions according to ecology or climate factors [
34
,
35
]. Our need was to characterize the basic climate drivers that influenced vegetation and recovery response, so instead of relying on these subdivisions, we derived gridded climate variables for each focal area directly. For each hurricane focal area, we derived the median monthly precipitation, maximum temperature, and vapor pressure deficit from 2000 to 2022 from the Climate Engine TerraClimate 4 km-Monthly dataset [
36
]. We then combined these climate time series for each focal area and analyzed their cross-seasonal and interannual similarities using non-hierarchical
k
-means clustering [
37
] using R 2024.04.2 software [
38
]. We explored prescribing different numbers of clusters to balance the need for representation with geographic coherence. The final iteration assigned a statistically-derived membership for each of our 44 hurricanes to six distinct climate subregions.
2.3. Hurricane Impacts and Their Drivers
High-frequency, multi-year LSP data broadly captures recurring seasonal behavior, hurricane impacts, and recovery. For each of the 44 focal areas, we calculated the areal mean 16-day EVI for a 22 year period (2001–2022) using 500 m MODIS Terra (MOD13A1) and Aqua (MYD13A1) data derived from Climate Engine (visited on July 2023). Instances of missing data from the source were handled using a non-parametric imputation method implemented in the missForest package in R 2024.04.2 software [
38
,
39
]. This algorithm employs Random Forests to estimate and replace missing values in a dataset. To document the normal seasonal LSP behavior for each cluster across years, we combined the 22 years of 16-day EVI values for the member hurricanes for an annual curve showing the mean and standard deviation. We then compared the dates of the hurricanes for each cluster with their annual LSP to clarify the implications of LSP seasonality for monitoring.
To understand impacts for each focal area, we used the MODIS Terra & Aqua 500 m 16-day dataset. We clipped the EVI time series to have two full years before and two years after each storm’s calendar year and derived two seasonal measures: the area under the curve (AUC) for the first winter (December through February) and the AUC for the first full year after the storm. Impacts were quantified as the percent change in EVI AUC (dEVI) from the mean of the two pre-storm seasonal values. For each subregional cluster, we used regression to compare measured impacts to the respective hurricane’s median wind speed that was derived from IBTrACS [
33
]. We then performed Mann–Whitney (Wilcoxon) tests on the regressions to compare differences among regions for the two seasonal measures.
Land cover varied among the focal areas and this could potentially influence the vulnerability of vegetation to the storms and our interpretation of short and long-term impacts. These patterns and differences among focal areas were assessed in Google Earth Engine using the MCD12Q1.061 MODIS Land Cover Type Yearly Global 500 m dataset. We summarized the percent land cover for each subregional cluster. To simplify categories across the region we reclassified the original classes into seven broader groups: mixed forest and woody savannas were merged into “mixed forest”; urban, water, wetland and other were grouped as “other”; and croplands, savannas, evergreen broadleaf forest and deciduous broadleaf forest were used as defined in the original dataset.
2.4. Local Case Studies
To achieve a more precise understanding of impacts for the six subregions, we analyzed the LSP on one representative category 4 hurricane for each subregional cluster. Hurricane Charley made landfall on 13 August 2004 near Cayo Costa, Florida, with a sustained wind of 210–249 km/h and caused an estimated
$
14 billion in economic losses. Hurricane Laura made landfall near Cameron, Louisiana on 27 August 2020 causing an estimated of
$
19 billion in economic losses. Hurricane Harvey, that was considered the second costliest U.S. tropical storm after Katrina, struck Texas on 26 August 2017 and caused an estimated damage of 125 billion. Hurricane Iota made landfall in Nicaragua on 17 November 2020, just two weeks after Eta passed over this same area. Hurricane Emily made landfall in the Yucatan Peninsula with sustained winds of 212 km/h. Hurricane Maria struck Puerto Rico on 20 September 2017 and resulted in severe damage. Detailed information on hurricane history and damage costs can be found at (
https://www.weather.gov/publications/assessments
, accessed on 24 February 2026,
https://www.noaa.gov/
, accessed on 24 February 2026).
For each of these six hurricanes, we derived a median EVI time series using the MODIS Terra (MOD13Q1) and Aqua (MYD13Q1) 250 m 16-day dataset and, respectively, that was compiled into an 8-day time series for all lands and forests alone. Our forest cover mask used the Global Land Cover ESA WorldCover 10 m v100 map (2020) dataset and was coarsened to 250 m. We smoothed each time series using a Savitsky–Golay algorithm in R 2024.04.2 software using a window of 23 observations to reduce noise [
38
,
40
,
41
,
42
]. Masking allowed us to resolve the potential influence of land cover on LSP for these six case studies and to help interpret results from the prior analysis of the 44 hurricanes. We calculated similar measures of impact as we had for the prior analysis. Percent change in EVI AUC was calculated as the difference between the mean of the two years prior to the hurricane compared to the year after the event. The EVI AUC of the first winter after the storm (December to February) was compared the mean of the prior two winters to minimize the effects of interannual variability.
3. Results
3.1. Climate Regionalization
Clustering grouped the 44 hurricanes into six types defined by climate, which we labeled according to their primary subregional location (
Figure 2
,
Table 1
). The Southeast U.S. (SUS) type had the most hurricanes at 13 and with all categories and a variety of wind speeds ranging from 148 to 259 km/h. The Caribbean (C) group included 10 hurricanes ranging from category 2 to 4. The Florida Peninsula (FP), South Gulf (SG) and Central America (CA), had six hurricanes each. The FP showed hurricanes categories 2 to 4 and the latter two from 2 to 5. With only three hurricanes represented, the West Gulf group had the fewest hurricanes, and these ranged from category 2 to 4. Clustering yielded strong spatial aggregations for FP, SUS, WG and CA. The location of C and SG were less well defined spatially, with the C group extending from Puerto Rico and Cuba to Belize and Mexico.
Land cover differed among the six subregions with respect to woody (deciduous, mixed, and evergreen forests) and other types (
Figure 3
). SG and CA averaged the most forest, with SUS and C averaging around 50% forest and FP and WG being predominantly savanna and grassland.
Subregions with the most forested land cover (SG and CA) show the highest median EVI while that with the least woody vegetation (WG) has the lowest EVI (
Figure 4
a). These differences in EVI are also consistent with the climate attributes of the subregions. CA’s high precipitation, warm temperature, and low VPD reflect its low latitude climate (
Figure 4
b–d). This contrasts with the least productive subregion, WG, that has the lowest precipitation and highest VPD (
Figure 4
b,d). SUS shows a broad range of EVI values and strong variation in temperature that reflects the strong seasonality of the higher-latitude climate and the large number of hurricanes included.
For all six subregions, hurricanes occurred near or following the peak of the growing season when vegetation was in normal seasonal decline (
Figure 5
). The median and standard deviation for each subregion shows the interannual variation over the 21-year period for the member hurricanes. Tropical CA had the highest year-round median EVI that consistently exceeded 0.5. SG had considerably more seasonal variability despite peaking with a higher EVI in July at the onset of its hurricane season. C shows a low median EVI below 0.4 early in the year, with the highest values around 0.55 during summer, indicating relatively moderate productivity with pronounced seasonality. Of all the subregions, C has the broadest range of hurricane dates represented, which is consistent with the Caribbean’s geographic exposure. FP and SUS show lower peak EVIs with strong seasonal variation, and WG exhibits the lowest median EVI values of all six subregions with a weak growing season peak that is comparatively stable between May and October when the three hurricanes occurred.
3.2. Hurricane Impacts
Of the six subregions, only SUS, FP and WG showed declines in EVI for both the first winter and first growing season after the storm, and the decline increased across those two periods (
Table 2
). While higher than expected for winter, C also shows a potential hurricane-associated decline for the first year after the storm. In contrast, more tropical CA showed the strongest mean decline of any subregion for the first winter, but none for the subsequent growing season. SG showed no measured decline on average, as EVI increased for both seasonal measures. Of all the subregions, SUS’s decline was most consistent across seasons.
Wind speed and EVI decline show an inconsistent relationship across subregions. For the first winter after the storms, stronger winds were associated with more decline for SUS (
p
= 0) and CA (
p
= 0.005), but other subregions showed no anticipated decline or pattern (
Figure 6
a). For the first year after the storm, SUS consistently showed a decline in impacts with increased wind speed (
p
= 0), but no other subregion showed this relationship (
Figure 6
b).
3.3. Local Case Studies
The six selected category 4 hurricanes show generally consistent results for all lands compared to their respective subregions (
Table 2
and
Table 3
). Five of the six hurricanes declined for the first winter, with Emily (SG) being the exception. Hurricane Maria (C) differed from its subregion’s response by also showing a first winter decline. For the first year after the storm, neither Maria (C) nor Charley (FP) showed a decline, but Iota (CA) did, although much less than in winter and unlike the response of the CA subregion overall. From the first winter to the first year, half of the cases show further decline while half improved, with Iota (CA) showing the most remarkable improvement.
The isolated forest responses of these six cases are directionally similar to their corresponding all-land responses but often with very different magnitudes (
Table 3
). All cases but Emily (SG) showed an initial drop in forest EVI, which was a directionally consistent but stronger increase than its all-lands response. Note that Emily occurred in mid-July, which was one to four months earlier than the dates of the other five cases (
Table 1
), and has extensive broadleaf forest cover (
Figure 3
). Iota (CA)’s forest decline was much stronger than its all-lands decline, but its EVI first-year recovery was greater. Maria (C) and Charley (FP) appear to have also recovered quickly. The forest EVI behavior of Iota, Maria and Charley is the opposite of what occurred with Laura (SUS), Harvey (WG) and Emily (SG); each of these showed increased EVI departure from the first winter to the first year after the storm.
Comparison of the five-year LSP behavior for forests and all lands shows a broadly similar form prior to the storms with Iota (CA) being most different (
Figure 7
). Hurricane impacts are less than the seasonal variation in EVI. There are often recurring differences in cover types in terms of their seasonal amplitude or peak timing. Note in particular the earlier forest EVI peak for Laura (SUS), Harvey (WG), and Emily (SG).
As LSP measures can vary among years for reasons other than disturbance such as seasonal weather, the EVI profiles for the two pre-hurricane years shown in
Figure 7
can inform the interpretation and attribution of change shown in
Table 3
. Note the more irregular two-year pre-storm LSP variability for Harvey (WG), Iota (CA), Emily (SG), and Maria (C) for the first winter or full year. Also note the more complex, less predictable LSP behavior of both forest and all lands for Iota (CA) over all five years.
4. Discussion
In this study, we investigated the impacts of 44 hurricanes that made landfall across six distinct subregions of the NAB using remote sensing. We show that seasonal insights from LSP help clarify and nuance our understanding of hurricane impacts, given the major climatic and ecological differences that are present across the region. Using two measures, the first winter after the storm and the first year after the storm, we demonstrated that higher wind speeds do not consistently correlate with greater storm impacts. For example, significant declines in EVI with increasing wind speed were observed only in the Southeast U.S., while other subregions showed weak or non-significant relationships (
Figure 6
). Phenologically, when hurricanes occurred during the mid growing season, as indicated on
Figure 5
, refoliation and rapid EVI recovery may have reduced our ability to recognize the magnitude of structural impacts. These ecological and technological limitations suggest the quality of remote sensing assessments using this common technique, particularly for tropical areas that have had the least field validations.
Isolating land cover types is a necessity. Non-woody vegetation types, such as grasslands or pasture, are biologically insensitive to wind damage, but they are often hypersensitive to unrelated climate variation that drives annual productivity. For analyses of change over time, the presence of these climate-sensitive types can distort the reliability of baseline conditions and the integrity of the change measure used to evaluate the post-storm vegetation state. This is a hazard to mixed land cover assessments when coarse-grid imagery is used, as non-targeted vegetation behavior can influence results. In heterogeneous landscapes, forest impacts may only be isolated through use of higher resolution imagery and masking that targets the susceptible vegetation. However, even when a high-resolution forest mask is available, it can be challenging to obtain high-resolution cloud-free imagery to conduct assessments at scale.
The phenological status of a forest at the time of the storm can affect the accuracy of impact assessments. An immediate response may indicate either structural tree damage or ephemeral leaf stripping, given the sensitivity of most spectral indices to foliage. Seasonally deciduous forest types may be particularly vulnerable to leaf stripping late in their growing season when their foliage is fragile due to senescence, and that is when most hurricanes occur (
Figure 5
). If the goal is to map structural damage in a rapid assessment, not just the general footprint of the storm more generally, leaf stripping can give false positives—a hazard that can be further aggravated with change detection approaches when the timing of senescence varies from year to year. To avoid this, we conducted our first impact assessment during the first winter rather than immediately after the storm (
Figure 5
), but this may have introduced a different problem as storms that occurred during the mid growing season may have experienced rapid refoliation prior to the winter assessment that persists. In the tropics, having a different time-since-storm could have affected the consistency of our first-winter measure among hurricanes. Remeasuring impacts across more than the two seasons used in this study may provide a way to isolate these complex phenological factors, but that can be difficult to achieve at scale in areas with frequent cloud cover. Very rapid assessments can also be misleading when hurricane rain provides drought relief through a short-lived resurgence of stressed vegetation, and this is a particularly confounding factor for areas with mixed land cover, pasture and grass. It is unclear if this rain response affected our results given the time of our winter assessment, but it could help explain surges of greenness reported by others using NDVI time series [
43
].
Crown and bole sprouting can initiate just weeks after a hurricane in tropical forests, and this phenomenon can obscure structural damage [
44
,
45
]. This phenomenon is less relevant for portions of SUS where hurricane damage to the predominant industrial loblolly pine forests often leads to tree mortality [
46
]. However, even north of SUS in the Appalachians, hardwood sprouting was prevalent after 2024’s hurricane Helene during the first growing season after the storm (personal observations of the second author, July 2025), suggesting that this is a widespread wind adaptive response. Sprouting vegetation can reduce the accuracy of impact assessments that use optical remote sensing, and this limits our understanding of forest resistance and recovery. Used alone or in combination, alternative technologies that capture change in vegetation height, such as high-resolution aerial LiDAR, largely bypass much of this problem as sprouting can occur on crown-damaged, leaning and fallen trees. Unfortunately, precise structural assessments using this technology take time and the extent of the analysis can be restricted [
47
].
Subregions
Our six subregions differ in terms of climate variability, land cover and vegetational sensitivity, and phenology. These attributes are ecologically inter-related and they provide useful ways to frame post-hurricane impact assessments. Awareness of these subregional attributes can help identify where research is lacking and help tailor post-hurricane monitoring to improve its effectiveness.
The Southeastern U.S. (SUS) stands out as a reliable area for long- and short-term monitoring of hurricane effects, and this is consistent with prior studies [
7
,
48
,
49
]. Impacts are resolvable during both winter and the subsequent growing season with impacts increasing at higher wind speeds, as expected (
Figure 6
). Other research has shown that damage to the industrial pine of this subregion is especially well-captured, as is the post-hurricane management response, but impacts to deciduous and mixed forests are harder to resolve [
18
]. Isolating leaf stripping from structural damage for deciduous forests can be challenging for storms that occur during fall senescence, but strong winter seasonality limits the potential for rapid hardwood sprouting.
In contrast, monitoring hurricane impacts in Central America (CA) is more challenging, particularly for the year after the storm. This is apparently due to the warm-wet tropical climate and high ecosystem productivity that causes rapid EVI recovery for the evergreen broadleaf and mixed forest types. For the first winter after the storm, we documented EVI declines that strengthened with wind speed, but we found no consistent evidence of impacts thereafter. If extreme hurricanes cause lasting structural damage to these forests, it is likely being obscured by refoliation, because over 70% of the land cover of CA is forest. Category 4 Iota had the strongest initial departure of any of the 44 hurricanes considered, but despite category 5 winds, Felix had less impact, although its impact was directionally similar to Iota over the two periods (
Table 1
,
Figure 6
). This may be because Felix’s focal area had substantially more grassland which is less sensitive to damage than forest (
Appendix A
,
Table A1
). The irregular year-to-year LSP shown for Hurricane Iota (
Figure 7
d) complicates assessment, and this may be caused by the quality of the remote sensing data given the frequency of clouds in this humid region. In CA, hurricanes tend to occur in late November, which is phenologically later than that other subregions (
Figure 5
). That difference in timing may affect the sensitivity of the deciduous canopy to defoliation or how aggressively hardwood refoliation occurs. Monitoring damage during first winter may be the most feasible for the CA subregion, but it is challenging to isolate structural impacts.
Only three of 44 hurricanes were included in the West Gulf (WG) subregion, and Harvey, the strongest, had the lowest mean wind speed of the six category 4 case studies (
Table 1
). Consistent with its warm-dry climate, WG had the lowest EVI across seasons and the least forest cover (
Figure 3
,
Figure 4
and
Figure 5
,
Appendix A
,
Table A1
). At less than 10% forest, the subregion’s vegetation should be inherently less sensitive to hurricane damage, given that grass is unlikely to retain damage into the next growing season, yet both forests and all lands show notable declines (
Table 2
and
Table 3
). It is particularly surprising that the first-year departure of all land cover is roughly double that of the forest for Harvey (
Table 3
). This suggests a cause from something other than the storm, such as the influence of interannual climate variability. Grass is particularly sensitive to climate variability and drought [
50
,
51
]. Our measure of storm impact, namely, change from a prior baseline, is particularly vulnerable to seasonal climate variability because either the pre- or post-disturbance value could be compromised. This problem appears to explain the first-year response of 2010’s category 2 Alex that, contrary to expectations, showed a stronger decline than the more intense storms for this subregion. Unlike the baseline used to calculate change, 2011’s first-year EVI was likely low given a severe drought as indicated by the North American Drought Monitor (
https://droughtmonitor.unl.edu/NADM/Maps.aspx
, accessed on 24 February 2026). This inherent limitation of monitoring in the subregion can reduce confidence in the accuracy of impact assessments, including those that rely on coarse-resolution forest masks. Results are also more challenging to interpret in this subregion with so few storms represented. For this subregion, forest monitoring requires a high spatial resolution analysis and sensitivity to interannual climate variability.
The South Gulf (SG) subregion is characterized by highly productive, seasonal forests with more than 80% forest cover, the highest of the six subregions (
Figure 3
). Annual EVI is the second-highest among the subregions, peaking as the highest but with a strong winter decline that shows high variability (
Figure 5
). Hurricanes in SG occurred from July, at peak EVI, through September, which is phenologically earlier than for those of other subregions. Rapid refoliation occurs during the growing season, and early storm dates may reduce the usefulness of first winter and first year measures (
Figure 6
). For example, Emily, the subregion’s case study, was the earliest of the six cases, and the fourth-earliest of all 44 storms analyzed (
Table 1
). This adaptation of tropical hardwoods may be more strongly expressed when defoliation occurs during the mid growing season, so impact assessment may be more accurate when conducted immediately after the hurricane occurs, although monitoring at this time can be difficult for low lying areas due to lingering floodwaters and the prevalence of non-structural defoliation.
The ten hurricanes analyzed for the Caribbean (C) subregion show varied responses for the first winter and first year after their storms with respect to wind speed. Some storms show a decline while others show an increase in EVI for these two periods. A recent study from Turner et al., [
30
], on the effects of hurricane Irma in Cuba found that while there was a significant decrease in EVI in mangroves and wetlands, there was a widespread increase in EVI elsewhere, including for the dry forest. In our analysis, Maria showed an EVI decline for the first winter after the storm, but an increase in EVI during the first year after the storm. This behavior is consistent with de Beurs et al., [
29
] who documented decline from Maria in Puerto Rico, with a quick recovery starting 8–12 weeks after the storm which they suggested might be associated with a growth surge from heavy rainfall. The timing of hurricanes here with respect to the region’s phenology suggests a strong potential for rapid refoliation and growth before the end of the growing season much of the time (
Figure 5
). In addition, with less forest cover than SG and CA and rugged topography that fragments forest cover, EVI’s sensitivity to post-storm weather variability may also be high. Similarly to SG, resolving hurricane impacts in this subregion can be challenging, but insights are possible soon after the storm. Use of high resolution imagery can help isolate the confounding responses of cover types.
Only about 10 percent of the Florida Peninsula (FP) subregion is forested, as savannas, grasslands, croplands, and other land cover types predominate. These latter types are less sensitive to damage, and this likely explains the non-responsiveness of EVI regardless of wind speed (
Figure 6
). The response of FP’s Hurricane Charley shows a short-term decline followed by rapid recovery for both all lands and forests, and this behavior is generally consistent with our detailed case studies for C and CA (
Table 3
). This may reflect the rapid sprouting of trees and shrubs that may or may not have substantial structural damage. While limited in distribution, hurricane damage to south Florida’s mangrove forests has received the most research attention and the homogeneity of the type makes it more easily monitored. The mangroves here have had four declines from severe hurricanes during the last three decades with slow recovery and a long-term decline [
52
]. Elsewhere, successful hurricane monitoring in the FP will likely benefit from high-resolution analyses that isolate the patchy woody plant cover.
5. Conclusions
This study identified inconsistencies in the predicted relationship between wind speed and hurricane damage. Subtropical SUS showed increased damage with wind speed—a pattern that remained across seasons. This anticipated relationship was absent elsewhere, as EVI change often increased after severe disturbance. Confounding drivers include subregional differences in land cover and climate-related productivity. Reduced evidence of storm impacts by the first growing season was common the tropics, which is consistent with higher productivity and rapid vegetative recovery, but some tropical hurricane footprints showed increased departure over time. This suggests that our measures of change may capture a combination of disturbance impacts and background variation in climate or LSP that could affect our ability to recognize change.
Our results underscore the need for a strategic approach to monitoring post-hurricane impacts that is tailored to the unique ecological conditions and land-use configurations of each subregion. Our findings emphasize the value of early monitoring for the tropics where rapid sprouting can obscure structural damage. However, rapid assessments that occur during the same growing season as the storm can misrepresent minor defoliation as structural damage. Rapid refoliation can occur in areas with or without structural damage, thus confounding estimates of vegetation structural recovery. This is not to say that rapid assessments with coarse spatial resolution lack value, as they can suggest the footprint of potential impacts where more focused evaluations can be pursued. This research shows that by the first year after the storm, structural impacts are difficult to resolve in the tropics, which means that subsequent monitoring of structural recovery will be even more difficult. This affects our ability to accurately assess the dynamics and resilience of these disturbance-prone ecosystems using this technology.
Our research shows that optical techniques are more effective for assessing immediate and long-term impacts to temperate regions such as SUS. This is not just because more research has been conducted there, as analyses benefit from the subregion’s relatively predictable phenological behavior and the sensitivity of industrial pines to storm damage [
18
]. The strong seasonality of temperate latitude forests means that rapid sprouting and refoliation are less likely to be concerns here. Outside this subregion, impact and monitoring assessments are more likely to struggle with distinguishing the recovery shown by spectral indices from structural recovery. The latter, more elusive phenomenon is needed to characterize long-term impacts, ecosystem resilience, and the global implications of these frequent, large-scale storms [
53
].
Hurricane impact monitoring will benefit from understanding how and why vegetation types vary from year to year normally. Non-woody vegetation may exhibit a surge in late-season growth from the influx of tropical storm rainfall in ways that interfere with rapid forest damage assessment, but it is unclear if this short-term mechanism affects the rate of woody plant refoliation. The cross-seasonal and year-to-year productivity of non-woody types can also vary in response to drought. Thus, weather variability can affect the integrity of either the normal baseline condition or the post-disturbance indicator of damage. Where vegetation is fragmented or mixed, monitoring at high spatial resolution with forest masks can reduce this problem. Conversely, hurricane footprints with relatively homogeneous forest cover have less issues with mixed responses from fractional cover and are more amenable to assessment from coarse resolution imagery.
Better use of LSP can improve our understanding of hurricane impacts across the ecologically and climatically diverse NAB. LSP encapsulates vegetation type and productivity, the seasonal and interannual dynamics that need consideration, and the actual storm impacts. LSP is more difficult to leverage in the humid tropics where high-resolution analyses, synthetic aperture radar (SAR) and (LiDAR) technologies that resolve structural change and recovery more precisely appear to be warranted.
Author Contributions
Conceptualization and methodology, S.P.N. and C.T.-P.; analysis and investigation, C.T.-P. and S.P.N.; data curation, C.T.-P. and S.P.N.; writing—original draft preparation, C.T.-P. and S.P.N.; writing—review and editing C.T.-P. and S.P.N.; visualization, C.T.-P. and S.P.N.; project administration: S.P.N. All authors have read and agreed to the published version of the manuscript.
Funding
This research received no external funding.
Data Availability Statement
The data supporting the conclusions of this article will be made available by the authors upon request.
Acknowledgments
This research was supported in part by an appointment to the United States Forest Service (USFS) Research Participation Program administered by the Oak Ridge Institute for Science and Education (ORISE) through an interagency agreement between the U.S. Department of Energy (DOE) and the U.S. Department of Agriculture (USDA). The findings and conclusions in this publication are those of the authors and should not be construed to represent any official USDA or U.S. Government determination or policy.
Conflicts of Interest
The authors declare that they have no commercial or associative interests that represent conflicts of interest in connection with the article submitted.
Abbreviations
The following abbreviations are used in this manuscript:
LSP
Land Surface Phenology
EVI
Enhanced Vegetation Index
AUC
Area Under the Curve
NAB
North Atlantic Basin
C
Caribbean
CA
Central America
FP
Florida Peninsula
SG
South Gulf
SUS
Southeast United States
WG
West Gulf
Appendix A
Table A1.
Percent of land cover types within the 44 hurricanes focal areas derived from MODIS.
Table A1.
Percent of land cover types within the 44 hurricanes focal areas derived from MODIS.
Florida Peninsula (FP)
Name
Croplands
Deciduous Broadleaf Forests
Evergreen Broadleaf Forests
Grasslands
Mixed Forests
Other
Savannas
FRANCES
1.1
0.4
0.4
6.7
1.2
8
82.2
IRMA
0.4
0.7
1.2
7.6
12.5
23.3
54.3
JEANNE
0.8
0.1
0.4
7.6
1.7
11.8
77.6
WILMA
1.9
0.5
3.2
3.9
18.7
34.1
37.8
CHARLEY *
0.2
0.1
0.7
12.4
2.9
9.1
74.6
IAN
0.1
0.1
0.8
9.9
3.1
9
77.1
Caribbean (C)
Name
Croplands
Deciduous Broadleaf Forests
Evergreen Broadleaf Forests
Grasslands
Mixed Forests
Other
Savannas
PALOMA
31.8
12
6.4
1.3
17.9
1.6
29
RICHARD
0.5
0
72.9
8.8
12.2
2.8
2.7
GRACE
8
0
5.7
2.6
50.9
2
30.9
IAN
22.4
0.1
17.3
5.7
21.9
5.4
27.2
SANDY
37.7
0.3
18.1
1.4
24.2
3.6
14.6
DENNIS
18.4
1.6
19.6
1.6
11.9
16.4
30.5
MARIA *
6.9
0
25
7.2
41.8
14.5
4.6
MICHELLE
31.6
7.7
13.1
0.6
9.5
6
31.5
WILMA
0
0.1
67.3
3.4
20.3
8.6
0.3
GUSTAV
27.9
0.2
15.6
3.3
16.5
5.3
31.3
South Gulf (SG)
Name
Croplands
Deciduous Broadleaf Forests
Evergreen Broadleaf Forests
Grasslands
Mixed Forests
Other
Savannas
DELTA
0.3
2
71.9
2.6
20.4
1.1
1.7
ERNESTO
0.1
0.1
80.5
2.9
11.7
3.8
0.9
ISIDORE
0.3
32.1
0
4.6
53.6
4
5.5
KARL
43.1
1.8
14.1
0.9
14.8
1.8
23.6
EMILY *
0.1
10.9
78.6
0.6
9.2
0.5
0
DEAN
0
0.2
81.3
2.4
11.7
3.7
0.7
Central America (CA)
Name
Croplands
Deciduous Broadleaf Forests
Evergreen Broadleaf Forests
Grasslands
Mixed Forests
Other
Savannas
BETA
0
0
29
1.9
33.5
3.7
31.8
IRIS
0
0
56.4
6.9
30.4
0.2
6
OTTO
0
0
52.7
1.4
25.7
0.9
19.3
ETA
0
0
43.6
9.8
33.5
8.5
4.6
IOTA *
0
0
45.2
4.2
32.7
6.2
11.7
FELIX
0
0
40.7
18.9
13
8.1
19.2
Southeast U.S. (SUS)
Name
Croplands
Deciduous Broadleaf Forests
Evergreen Broadleaf Forests
Grasslands
Mixed Forests
Other
Savannas
DELTA
39.4
3.4
2.4
4.7
8.7
15
26.5
GUSTAV
2.1
6
0.1
1.9
6.2
75.3
8.4
IKE
0.9
2.6
8.7
6.4
39.4
16.4
25.7
ISABEL
21.3
6.2
3.1
1.6
55.9
9.9
2.1
SALLY
10.8
3.1
14.4
7.6
51.1
3
10
DENNIS
5.7
11.3
10.4
3.8
62.1
2.6
4.1
IVAN
4.5
19.4
12.8
3
47.4
7.6
5.3
KATRINA
0
11
6.8
2.2
66.7
5.5
7.8
RITA
0.2
5.8
11.5
2.3
40.6
17.9
21.7
ZETA
1.3
2.6
0
2.7
4.2
62.6
26.5
IDA
8.6
15.7
0
3
12
39
21.7
LAURA *
4.1
2
13.4
4.4
32.4
19.1
24.6
MICHAEL
4.9
5.5
22.5
3.6
52.4
6
5.1
West Gulf (WG)
Name
Croplands
Deciduous Broadleaf Forests
Evergreen Broadleaf Forests
Grasslands
Mixed Forests
Other
Savannas
ALEX
0.4
4.7
2.5
44.5
6
5.9
35.9
EMILY
5.1
0.1
0
66.3
2.4
15.7
10.3
HARVEY *
4.4
0.1
0
25.3
0
4.6
65.7
* The hurricanes examined as case studies.
References
Cole, J.; Nowak, D.J.; Greenfield, E.J. Potential Hurricane Wind Risk to US Rural and Urban Forests.
J. For.
2021
,
119
, 393–406. [
Google Scholar
] [
CrossRef
]
Mallin, M.A.; Corbett, C.A. How Hurricane Attributes Determine the Extent of Environmental Effects: Multiple Hurricanes and Different Coastal Systems.
Estuaries Coasts
2006
,
29
, 1046–1061. [
Google Scholar
] [
CrossRef
]
Reed, K.A.; Wehner, M.F.; Zarzycki, C.M. Attribution of 2020 Hurricane Season Extreme Rainfall to Human-Induced Climate Change.
Nat. Commun.
2022
,
13
, 1905. [
Google Scholar
] [
CrossRef
] [
PubMed
]
Sharma, A.; Ojha, S.K.; Dimov, L.D.; Vogel, J.G.; Nowak, J. Long-Term Effects of Catastrophic Wind on Southern US Coastal Forests: Lessons from a Major Hurricane.
PLoS ONE
2021
,
16
, e0243362. [
Google Scholar
] [
CrossRef
]
Marsooli, R.; Lin, N.; Emanuel, K.; Feng, K. Climate Change Exacerbates Hurricane Flood Hazards along US Atlantic and Gulf Coasts in Spatially Varying Patterns.
Nat. Commun.
2019
,
10
, 3785. [
Google Scholar
] [
CrossRef
] [
PubMed
]
Vecchi, G.A.; Landsea, C.; Zhang, W.; Villarini, G.; Knutson, T. Changes in Atlantic Major Hurricane Frequency since the Late-19th Century.
Nat. Commun.
2021
,
12
, 4054. [
Google Scholar
] [
CrossRef
] [
PubMed
]
Lin, T.C.; Hogan, J.A.; Chang, C.T. Tropical Cyclone Ecology: A Scale-Link Perspective.
Trends Ecol. Evol.
2020
,
35
, 594–604. [
Google Scholar
] [
CrossRef
]
Gao, W.; Larjavaara, M. Wind Disturbance in Forests: A Bibliometric Analysis and Systematic Review.
For. Ecol. Manag.
2024
,
564
, 122001. [
Google Scholar
] [
CrossRef
]
Hernandez, J.O.; Maldia, L.S.J.; Park, B.B.; Hernandez, J.O.; Maldia, L.S.J.; Park, B.B. Research Trends and Methodological Approaches of the Impacts of Windstorms on Forests in Tropical, Subtropical, and Temperate Zones: Where Are We Now and How Should Research Move Forward?
Plants
2020
,
9
, 1709. [
Google Scholar
] [
CrossRef
] [
PubMed
]
Henebry, G.M.; de Beurs, K.M. Remote Sensing of Land Surface Phenology: A Prospectus. In
Phenology: An Integrative Environmental Science
; Schwartz, M.D., Ed.; Springer: Dordrecht, The Netherlands, 2013; pp. 385–411. [
Google Scholar
]
Pomara, L.Y.; Lee, D.C.; Brooks, B.-G.; Hargrove, W.W. Using Land Surface Phenology and Information Theory to Assess and Map Complex Landscape Dynamics.
Landsc. Ecol.
2024
,
39
, 203. [
Google Scholar
] [
CrossRef
]
White, M.A.; de Beurs, K.M.; Didan, K.; Inouye, D.W.; Richardson, A.D.; Jensen, O.P.; O’Keefe, J.; Zhang, G.; Nemani, R.R.; van Leeuwen, W.J.D.; et al. Intercomparison, Interpretation, and Assessment of Spring Phenology in North America Estimated from Remote Sensing for 1982-2006.
Glob. Change Biol.
2009
,
15
, 2335–2359. [
Google Scholar
] [
CrossRef
]
Caparros-Santiago, J.A.; Rodriguez-Galiano, V.; Dash, J. Land Surface Phenology as Indicator of Global Terrestrial Ecosystem Dynamics: A Systematic Review.
ISPRS J. Photogramm. Remote Sens.
2021
,
171
, 330–347. [
Google Scholar
] [
CrossRef
]
Kooistra, L.; Berger, K.; Brede, B.; Graf, L.V.; Aasen, H.; Roujean, J.L.; Machwitz, M.; Schlerf, M.; Atzberger, C.; Prikaziuk, E.; et al. Reviews and Syntheses: Remotely Sensed Optical Time Series for Monitoring Vegetation Productivity.
Biogeosciences
2024
,
21
, 473–511. [
Google Scholar
] [
CrossRef
]
Cui, K.L.; Yang, J.L.; Dong, J.W.; Zhao, G.S.; Cui, Y.P. Comparing Different Spatial Resolutions and Indices for Retrieving Land Surface Phenology for Deciduous Broadleaf Forests.
Remote Sens.
2023
,
15
, 2266. [
Google Scholar
] [
CrossRef
]
Gnilke, A.; Sanders, T.G.M. Distinguishing Abrupt and Gradual Forest Disturbances with MODIS-Based Phenological Anomaly Series.
Front. Plant Sci.
2022
,
13
, 863116. [
Google Scholar
] [
CrossRef
]
Guzmán Q., J.A.; Pinto-Ledezma, J.N.; Frantz, D.; Townsend, P.A.; Juzwik, J.; Cavender-Bares, J. Mapping Oak Wilt Disease from Space Using Land Surface Phenology.
Remote Sens. Environ.
2023
,
298
, 113794. [
Google Scholar
] [
CrossRef
]
Topete-Pozas, C.; Norman, S.P.; Christie, W.M. Multi-Year Hurricane Impacts Across an Urban-to-Industrial Forest Use Gradient.
Remote Sens.
2024
,
16
, 3890. [
Google Scholar
] [
CrossRef
]
Wang, F.G.; Xu, Y.J. Hurricane Katrina-Induced Forest Damage in Relation to Ecological Factors at Landscape Scale.
Environ. Monit. Assess.
2009
,
156
, 491–507. [
Google Scholar
] [
CrossRef
]
Wang, W.T.; Qu, J.J.; Hao, X.J.; Liu, Y.Q.; Stanturf, J.A. Post-Hurricane Forest Damage Assessment Using Satellite Remote Sensing.
Agric. For. Meteorol.
2010
,
150
, 122–132. [
Google Scholar
] [
CrossRef
]
Feng, Y.L.; Negron-Juarez, R.I.; Chambers, J.Q. Remote Sensing and Statistical Analysis of the Effects of Hurricane Maria on the Forests of Puerto Rico.
Remote Sens. Environ.
2020
,
247
, 111940. [
Google Scholar
] [
CrossRef
]
Dupuis, C.; Lejeune, P.; Michez, A.; Fayolle, A. How Can Remote Sensing Help Monitor Tropical Moist Forest Degradation?—A Systematic Review.
Remote Sens.
2020
,
12
, 1087. [
Google Scholar
] [
CrossRef
]
Helman, D. Land Surface Phenology: What Do We Really “see” from Space?
Sci. Total Environ.
2018
,
618
, 665–673. [
Google Scholar
] [
CrossRef
]
Huete, A.; Didan, K.; Miura, T.; Rodriguez, E.P.; Gao, X.; Ferreira, L.G. Overview of the Radiometric and Biophysical Performance of the MODIS Vegetation Indices.
Remote Sens. Environ.
2002
,
83
, 195–213. [
Google Scholar
] [
CrossRef
]
Rogan, J.; Schneider, L.; Christman, Z.; Millones, M.; Lawrence, D.; Schmook, B. Hurricane Disturbance Mapping Using MODIS EVI Data in the Southeastern Yucatan, Mexico.
Remote Sens. Lett.
2011
,
2
, 259–267. [
Google Scholar
] [
CrossRef
]
Rossi, E.; Rogan, J.; Schneider, L. Mapping Forest Damage in Northern Nicaragua after Hurricane Felix (2007) Using MODIS Enhanced Vegetation Index Data.
Gisci. Remote Sens.
2013
,
50
, 385–399. [
Google Scholar
] [
CrossRef
]
Wang, F.G.; D’Sa, E.J. Potential of MODIS EVI in Identifying Hurricane Disturbance to Coastal Vegetation in the Northern Gulf of Mexico.
Remote Sens.
2010
,
2
, 1–18. [
Google Scholar
] [
CrossRef
]
Gang, C.C.; Pan, S.F.; Tian, H.Q.; Wang, Z.N.; Xu, R.T.; Bian, Z.H.; Pan, N.Q.; Yao, Y.Z.; Shi, H. Satellite Observations of Forest Resilience to Hurricanes along the Northern Gulf of Mexico.
For. Ecol. Manag.
2020
,
472
, 118243. [
Google Scholar
] [
CrossRef
]
de Beurs, K.M.; McThompson, N.S.; Owsley, B.C.; Henebry, G.M. Hurricane Damage Detection on Four Major Caribbean Islands.
Remote Sens. Environ.
2019
,
229
, 1–13. [
Google Scholar
] [
CrossRef
]
Turner, H.C.; Galford, G.L.; Lopez, N.H.; Méndez, A.F.; Borroto-Escuela, D.Y.; Ramos, I.H.; González-Díaz, P. Extent, Severity, and Temporal Patterns of Damage to Cuba’s Ecosystems Following Hurricane Irma: MODIS and Sentinel-2 Hurricane Disturbance Vegetation Anomaly (HDVA).
Remote Sens.
2023
,
15
, 2495. [
Google Scholar
] [
CrossRef
]
Dronova, I.; Taddeo, S. Remote Sensing of Phenology: Towards the Comprehensive Indicators of Plant Community Dynamics from Species to Regional Scales.
J. Ecol.
2022
,
110
, 1460–1484. [
Google Scholar
] [
CrossRef
]
Hogan, J.A.; Feagin, R.A.; Starr, G.; Ross, M.; Lin, T.C.; O’connell, C.; Huff, T.P.; Stauffer, B.A.; Robinson, K.L.; Lara, M.C.; et al. A Research Framework to Integrate Cross-Ecosystem Responses to Tropical Cyclones.
Bioscience
2020
,
70
, 477–489. [
Google Scholar
] [
CrossRef
]
Kenneth, R.; Howard, J.; James, P.; Michael, C.; Carl, J.
International Best Track Archive for Climate Stewardship (IBTrACS) Project
; Version 4; NOAA National Centers for Environmental Information (NCEI): Asheville, NC, USA, 2019. [
Google Scholar
]
Zhou, Y.; Matyas, C.J. Regionalization of Precipitation Associated with Tropical Cyclones Using Spatial Metrics and Satellite Precipitation.
GISci. Remote Sens.
2021
,
58
, 542–561. [
Google Scholar
] [
CrossRef
]
Kim, D.; Matyas, C.J. Classification of Tropical Cyclone Rain Patterns Using Convolutional Autoencoder.
Sci. Rep.
2024
,
14
, 791. [
Google Scholar
] [
CrossRef
]
Abatzoglou, J.T.; Dobrowski, S.Z.; Parks, S.A.; Hegewisch, K.C. TerraClimate, a High-Resolution Global Dataset of Monthly Climate and Climatic Water Balance from 1958–2015.
Sci. Data
2018
,
5
, 170191. [
Google Scholar
] [
CrossRef
] [
PubMed
]
Hargrove, W.W.; Hoffman, F.M. Using Multivariate Clustering to Characterize Ecoregion Borders.
Comput. Sci. Eng.
1999
,
1
, 18–25. [
Google Scholar
] [
CrossRef
]
R Core Team.
R: A Language and Environment for Statistical Computing
; R Core Team: Vienna, Austria, 2024. [
Google Scholar
]
Stekhoven, D.J.
missForest: Nonparametric Missing Value Imputation Using Random Forest
; Harvard University: Cambridge, MA, USA, 2025. [
Google Scholar
]
Kong, D.D.; Zhang, Y.Q.; Gu, X.H.; Wang, D.G. A Robust Method for Reconstructing Global MODIS EVI Time Series on the Google Earth Engine.
ISPRS J. Photogramm. Remote Sens.
2019
,
155
, 13–24. [
Google Scholar
] [
CrossRef
]
Savitzky, A.; Golay, M.J.E. Smoothing + Differentiation of Data by Simplified Least Squares Procedures.
Anal. Chem.
1964
,
36
, 1627–1639. [
Google Scholar
] [
CrossRef
]
Tian, J.Q.; Zhu, X.L.; Chen, J.; Wang, C.; Shen, M.G.; Yang, W.; Tan, X.Y.; Xu, S.; Li, Z.L. Improving the Accuracy of Spring Phenology Detection by Optimally Smoothing Satellite Vegetation Index Time Series Based on Local Cloud Frequency.
ISPRS J. Photogramm. Remote Sens.
2021
,
180
, 29–44. [
Google Scholar
] [
CrossRef
]
Sánchez-Rivera, G.; Gómez-Mendoza, L. Tropical Cyclone Effects on Vegetation Resilience in the Yucatan Peninsula, México, between 2000–2012.
Investig. Geográficas
2022
,
203
, 203–220. [
Google Scholar
] [
CrossRef
]
Basnet, K. Recovery of a Tropical Rain Forest after Hurricane Damage.
Vegetatio
1993
,
109
, 1–4. [
Google Scholar
] [
CrossRef
]
Taylor, B.N.; Stedman, E.; Van Bloem, S.J.; Whitmire, S.L.; DeWalt, S.J. Widespread Stem Snapping but Limited Mortality Caused by a Category 5 Hurricane on the Caribbean Island of Dominica.
For. Ecol. Manag.
2023
,
532
, 120833. [
Google Scholar
] [
CrossRef
]
Johnsen, K.H.; Butnor, J.R.; Kush, J.S.; Schmidtling, R.C.; Nelson, C.D. Hurricane Katrina Winds Damaged Longleaf Pine Less than Loblolly Pine.
South. J. Appl. For.
2009
,
33
, 178–181. [
Google Scholar
] [
CrossRef
]
Leitold, V.; Morton, D.C.; Martinuzzi, S.; Paynter, I.; Uriarte, M.; Keller, M.; Ferraz, A.; Cook, B.D.; Corp, L.A.; González, G. Tracking the Rates and Mechanisms of Canopy Damage and Recovery Following Hurricane Maria Using Multitemporal Lidar Data.
Ecosystems
2022
,
25
, 892–910. [
Google Scholar
] [
CrossRef
]
Amaral, C.; Poulter, B.; Lagomasino, D.; Fatoyinbo, T.; Taillie, P.; Lizcano, G.; Canty, S.; Silveira, J.A.H.; Teutli-Hernández, C.; Cifuentes-Jara, M.; et al. Drivers of Mangrove Vulnerability and Resilience to Tropical Cyclones in the North Atlantic Basin.
Sci. Total Environ.
2023
,
898
, 165413. [
Google Scholar
] [
CrossRef
]
Gong, Y.; Staudhammer, C.L.; Kenney, G.; Wiesner, S.; Zhang, Y.L.; Starr, G. Vegetation Structure Drives Forest Phenological Recovery after Hurricane.
Sci. Total Environ.
2021
,
774
, 145651. [
Google Scholar
] [
CrossRef
]
Munson, S.M.; Long, A.L. Climate Drives Shifts in Grass Reproductive Phenology across the Western USA.
New Phytol.
2017
,
213
, 1945–1955. [
Google Scholar
] [
CrossRef
] [
PubMed
]
Norman, S.P.; Koch, F.H.; Hargrove, W.W. Review of Broad-Scale Drought Monitoring of Forests: Toward an Integrated Data Mining Approach.
For. Ecol. Manag.
2016
,
380
, 346–358. [
Google Scholar
] [
CrossRef
]
Liang, X.; Dai, Z.; Mei, X.; Wang, R.; Zeng, W.; Fagherazzi, S. Hurricanes Induced Irreversible Large-Scale Loss of Mangrove Forests.
Geophys. Res. Lett.
2025
,
52
, e2025GL115692. [
Google Scholar
] [
CrossRef
]
Dahal, D.; Liu, S.; Oeding, J. The Carbon Cycle and Hurricanes in the United States between 1900 and 2011.
Sci. Rep.
2014
,
4
, 5197. [
Google Scholar
] [
CrossRef
]
Figure 1.
Flowchart showing the data sources (in gray) and methodological workflow.
Figure 1.
Flowchart showing the data sources (in gray) and methodological workflow.
Figure 2.
The NAB study area showing the 44 hurricane tracks. The rectangles are the focal areas used in the analyses and the colors show the results of clustering.
Figure 2.
The NAB study area showing the 44 hurricane tracks. The rectangles are the focal areas used in the analyses and the colors show the results of clustering.
Figure 3.
The fraction of land cover types for the combined hurricane focal areas for each subregion. For detailed land cover by hurricane, see
Table A1
. Region labels are the same as in
Figure 2
.
Figure 3.
The fraction of land cover types for the combined hurricane focal areas for each subregion. For detailed land cover by hurricane, see
Table A1
. Region labels are the same as in
Figure 2
.
Figure 4.
Regional inter-seasonal and interannual vegetation and climate variation over the full period of record 2001–2022 for the hurricanes included in each subregion. The bold line shows the median value and the boxes show central distribution of the data: (
a
) EVI, (
b
) precipitation, (
c
) temperature, (
d
) vapor pressure deficit.
Figure 4.
Regional inter-seasonal and interannual vegetation and climate variation over the full period of record 2001–2022 for the hurricanes included in each subregion. The bold line shows the median value and the boxes show central distribution of the data: (
a
) EVI, (
b
) precipitation, (
c
) temperature, (
d
) vapor pressure deficit.
Figure 5.
The timing of hurricanes with respect to annual land surface phenology (LSP). The curved lines show the 22-year median and first standard deviations of EVI for the hurricane focal areas included in each subregion. Vertical lines show the landfall dates for the hurricanes included in each subregion. The colors are the same as used in
Figure 2
.
Figure 5.
The timing of hurricanes with respect to annual land surface phenology (LSP). The curved lines show the 22-year median and first standard deviations of EVI for the hurricane focal areas included in each subregion. Vertical lines show the landfall dates for the hurricanes included in each subregion. The colors are the same as used in
Figure 2
.
Figure 6.
Percent change in EVI AUC compared to the median wind speed of the hurricane’s focal areas by subregion for (
a
) the first winter after the storm, and (
b
) the first year after the storm. The colors are the same as used in
Figure 2
.
Figure 6.
Percent change in EVI AUC compared to the median wind speed of the hurricane’s focal areas by subregion for (
a
) the first winter after the storm, and (
b
) the first year after the storm. The colors are the same as used in
Figure 2
.
Figure 7.
Five-year forest (green) and all lands (gray) LSP median EVI profiles for the six case studies centered on the year of the hurricane for (
a
) Charley, (
b
) Laura, (
c
) Harvey, (
d
) Iota, (
e
) Emily, and (
f
) Maria. Landfall dates are shown by red dashed lines.
Figure 7.
Five-year forest (green) and all lands (gray) LSP median EVI profiles for the six case studies centered on the year of the hurricane for (
a
) Charley, (
b
) Laura, (
c
) Harvey, (
d
) Iota, (
e
) Emily, and (
f
) Maria. Landfall dates are shown by red dashed lines.
Table 1.
The 44 hurricanes analyzed in this study by subregion. The six storms selected for in-depth analysis as case studies are denoted with an *. The wind speed corresponds to the median value within the respective track/focal area.
Table 1.
The 44 hurricanes analyzed in this study by subregion. The six storms selected for in-depth analysis as case studies are denoted with an *. The wind speed corresponds to the median value within the respective track/focal area.
Florida Peninsula (FP)
Central America (CA)
Name
Category
Wind speed (km/h)
Landfall
Name
Category
Wind speed (km/h)
Landfall
FRANCES
2
157.42
2 September 2004
BETA
2
133.34
30 October 2005
IRMA
3
178.71
6 September 2017
IRIS
3
200
9 October 2001
JEANNE
3
185.2
16 September 2004
OTTO
3
180.57
24 November 2016
WILMA
3
185.2
24 October 2005
ETA
4
175.94
3 November 2020
CHARLEY *
4
231.5
13 August 2004
IOTA *
4
203.72
17 November 2020
IAN
4
207.42
28 September 2022
FELIX
5
233.35
4 September 2007
Caribbean (C)
Southeast U.S. (SUS)
Name
Category
Wind speed (km/h)
Landfall
Name
Category
Wind speed (km/h)
Landfall
PALOMA
2
118.52
8 November 2008
DELTA
2
157.42
9 October 2020
RICHARD
2
148.16
25 October 2010
GUSTAV
2
171.31
1 September 2008
GRACE
3
185.2
19 August 2021
IKE
2
162.97
7 September 2008
IAN
3
197.23
27 September 2022
ISABEL
2
148.16
18 September 2003
SANDY
3
185.2
24 October 2012
SALLY
2
166.68
16 September 2020
DENNIS
4
213.9
8 July 2005
DENNIS
3
175.94
10 July 2005
MARIA *
4
212.98
20 September 2017
IVAN
3
180.57
16 September 2004
MICHELLE
4
203.72
4 November 2001
KATRINA
3
194.46
25 August 2005
WILMA
4
185.2
22 October 2005
RITA
3
175.01
24 September 2005
GUSTAV
4
236.13
30 August 2008
ZETA
3
185.2
27 October 2020
IDA
4
212.98
29 August 2021
LAURA *
4
219.46
27 August 2020
MICHAEL
5
259.28
10 October 2018
South Gulf (SG)
West Gulf (WG)
Name
Category
Wind speed (km/h)
Landfall
Name
Category
Wind speed (km/h)
Landfall
DELTA
2
166.68
7 October 2020
ALEX
2
166.68
1 July 2010
ERNESTO
2
138.9
8 August 2012
EMILY
3
185.2
20 July 2005
ISIDORE
3
185.2
22 September 2002
HARVEY *
4
194.46
26 August 2017
KARL
3
175.94
17 September 2010
EMILY *
4
212.98
18 July 2005
DEAN
5
235.2
21 August 2007
Table 2.
Mean percent change in EVI AUC by subregion.
Table 2.
Mean percent change in EVI AUC by subregion.
Subregion
1st Winter
1st Year
Florida Peninsula (FP)
−0.02
−1.25
Central America (CA)
−9.69
0.3
Caribbean (C)
0.39
−1.28
Southeast U.S. (SUS)
−2.58
−2.78
South Gulf (SG)
2.33
0.1
West Gulf (WG)
−0.72
−9.23
Table 3.
Immediate and secondary response of EVI AUC for forest and all lands for the six hurricane case studies. Units are percent change in the sum of the 16-day EVI values under the winter or annual LSP curve.
Table 3.
Immediate and secondary response of EVI AUC for forest and all lands for the six hurricane case studies. Units are percent change in the sum of the 16-day EVI values under the winter or annual LSP curve.
Cases
All Lands
Forest
1st Winter
1st Year
1st Winter
1st Year
Charley (FP)
−2.32
2.74
−2.47
1.17
Laura (SUS)
−7.88
−9.43
−10.23
−13.64
Harvey (WG)
−9.47
−9.83
−2.25
−4.72
Iota (CA)
−23.53
−5.17
−35.84
−2.96
Emily (SG)
6.44
−6.63
11.4
−4.9
Maria (C)
−7.85
0.9
−1.12
1.64
Disclaimer/Publisher’s Note:
The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the
Creative Commons Attribution (CC BY) license
.
Share and Cite
MDPI and ACS Style
Topete-Pozas, C.; Norman, S.P.
Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022).
Forests
2026
,
17
, 419.
https://doi.org/10.3390/f17040419
AMA Style
Topete-Pozas C, Norman SP.
Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022).
Forests
. 2026; 17(4):419.
https://doi.org/10.3390/f17040419
Chicago/Turabian Style
Topete-Pozas, Carlos, and Steven P. Norman.
2026. "Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022)"
Forests
17, no. 4: 419.
https://doi.org/10.3390/f17040419
APA Style
Topete-Pozas, C., & Norman, S. P.
(2026). Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022).
Forests
,
17
(4), 419.
https://doi.org/10.3390/f17040419
Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details
here
.
Article Metrics |
| Markdown | Next Article in Journal
[Biomass Estimation of *Picea schrenkiana* Forests in the Western Tianshan Mountains Using Integrated ICESat-2 and GF-6 Data](https://www.mdpi.com/1999-4907/17/4/421)
Previous Article in Journal
[Drought Stress Response of Norway Spruce Seedlings Treated with Drought-Mitigative Additives](https://www.mdpi.com/1999-4907/17/4/420)
Previous Article in Special Issue
[Generalist Pests Cause High Tree Infestation, but Specialist Pests Cause High Mortality](https://www.mdpi.com/1999-4907/16/1/127)
## Journals
[Active Journals](https://www.mdpi.com/about/journals) [Find a Journal](https://www.mdpi.com/about/journalfinder) [Journal Proposal](https://www.mdpi.com/about/journals/proposal) [Proceedings Series](https://www.mdpi.com/about/proceedings)
[Topics](https://www.mdpi.com/topics)
## Information
[For Authors](https://www.mdpi.com/authors) [For Reviewers](https://www.mdpi.com/reviewers) [For Editors](https://www.mdpi.com/editors) [For Librarians](https://www.mdpi.com/librarians) [For Publishers](https://www.mdpi.com/publishing_services) [For Societies](https://www.mdpi.com/societies) [For Conference Organizers](https://www.mdpi.com/conference_organizers)
[Open Access Policy](https://www.mdpi.com/openaccess) [Institutional Open Access Program](https://www.mdpi.com/ioap) [Special Issues Guidelines](https://www.mdpi.com/special_issues_guidelines) [Editorial Process](https://www.mdpi.com/editorial_process) [Research and Publication Ethics](https://www.mdpi.com/ethics) [Article Processing Charges](https://www.mdpi.com/apc) [Awards](https://www.mdpi.com/awards) [Testimonials](https://www.mdpi.com/testimonials)
[Author Services](https://www.mdpi.com/authors/english)
## Initiatives
[Sciforum](https://sciforum.net/) [MDPI Books](https://www.mdpi.com/books) [Preprints.org](https://www.preprints.org/) [Scilit](https://www.scilit.com/) [SciProfiles](https://sciprofiles.com/) [Encyclopedia](https://encyclopedia.pub/) [JAMS](https://jams.pub/) [Proceedings Series](https://www.mdpi.com/about/proceedings)
## About
[Overview](https://www.mdpi.com/about) [Contact](https://www.mdpi.com/about/contact) [Careers](https://careers.mdpi.com/) [News](https://www.mdpi.com/about/announcements) [Press](https://www.mdpi.com/about/press) [Blog](http://blog.mdpi.com/)
[Sign In / Sign Up](https://www.mdpi.com/user/login)
## Notice
[*clear*]()
## Notice
You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.
[Continue]() [Cancel]()
[*clear*]()
All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to <https://www.mdpi.com/openaccess>.
Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.
Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.
Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.
Original Submission Date Received: .
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
[](https://www.mdpi.com/)
[*clear*](https://www.mdpi.com/1999-4907/17/4/419) [*zoom\_out\_map*](https://www.mdpi.com/toggle_desktop_layout_cookie "Toggle desktop layout") [*search*](https://www.mdpi.com/1999-4907/17/4/419) [*menu*](https://www.mdpi.com/1999-4907/17/4/419 "MDPI main page")
[](https://www.mdpi.com/)
- [Journals](https://www.mdpi.com/about/journals)
- - [Active Journals](https://www.mdpi.com/about/journals)
- [Find a Journal](https://www.mdpi.com/about/journalfinder)
- [Journal Proposal](https://www.mdpi.com/about/journals/proposal)
- [Proceedings Series](https://www.mdpi.com/about/proceedings)
- [Topics](https://www.mdpi.com/topics)
- [Information](https://www.mdpi.com/authors)
- - [For Authors](https://www.mdpi.com/authors)
- [For Reviewers](https://www.mdpi.com/reviewers)
- [For Editors](https://www.mdpi.com/editors)
- [For Librarians](https://www.mdpi.com/librarians)
- [For Publishers](https://www.mdpi.com/publishing_services)
- [For Societies](https://www.mdpi.com/societies)
- [For Conference Organizers](https://www.mdpi.com/conference_organizers)
- [Open Access Policy](https://www.mdpi.com/openaccess)
- [Institutional Open Access Program](https://www.mdpi.com/ioap)
- [Special Issues Guidelines](https://www.mdpi.com/special_issues_guidelines)
- [Editorial Process](https://www.mdpi.com/editorial_process)
- [Research and Publication Ethics](https://www.mdpi.com/ethics)
- [Article Processing Charges](https://www.mdpi.com/apc)
- [Awards](https://www.mdpi.com/awards)
- [Testimonials](https://www.mdpi.com/testimonials)
- [Author Services](https://www.mdpi.com/authors/english)
- [Initiatives](https://www.mdpi.com/about/initiatives)
- - [Sciforum](https://sciforum.net/)
- [MDPI Books](https://www.mdpi.com/books)
- [Preprints.org](https://www.preprints.org/)
- [Scilit](https://www.scilit.com/)
- [SciProfiles](https://sciprofiles.com/)
- [Encyclopedia](https://encyclopedia.pub/)
- [JAMS](https://jams.pub/)
- [Proceedings Series](https://www.mdpi.com/about/proceedings)
- [About](https://www.mdpi.com/about)
- - [Overview](https://www.mdpi.com/about)
- [Contact](https://www.mdpi.com/about/contact)
- [Careers](https://careers.mdpi.com/)
- [News](https://www.mdpi.com/about/announcements)
- [Press](https://www.mdpi.com/about/press)
- [Blog](http://blog.mdpi.com/)
[Sign In / Sign Up](https://www.mdpi.com/user/login) [Submit](https://www.mdpi.com/1999-4907/17/4/%20%20%20%20%20%20%20%20%20%20%20%20https://susy.mdpi.com/user/manuscripts/upload?journal=forests%0A%20%20%20%20)
*error\_outline* You can access [the new MDPI.com website here](https://www.mdpi.com/redirect/new_site?return=/1999-4907/17/4/419). Explore and share your feedback with us. [*close*](https://www.mdpi.com/1999-4907/17/4/419)
[Journals](https://www.mdpi.com/about/journals)
[Forests](https://www.mdpi.com/journal/forests)
[Volume 17](https://www.mdpi.com/1999-4907/17)
[Issue 4](https://www.mdpi.com/1999-4907/17/4)
[10\.3390/f17040419](https://www.mdpi.com/1999-4907/17/4/419)
[](https://www.mdpi.com/journal/forests)
[Submit to this Journal](https://susy.mdpi.com/user/manuscripts/upload?form%5Bjournal_id%5D%3D42) [Review for this Journal](https://susy.mdpi.com/volunteer/journals/review) [Propose a Special Issue](https://www.mdpi.com/journalproposal/sendproposalspecialissue/forests)
[► ▼ Article Menu](https://www.mdpi.com/1999-4907/17/4/419)
## Article Menu
- [Academic Editors](https://www.mdpi.com/1999-4907/17/4/419#academic_editors)
[Baozhang Chen](https://sciprofiles.com/profile/67643?utm_source=mdpi.com&utm_medium=website&utm_campaign=avatar_name)
[Romà Ogaya](https://sciprofiles.com/profile/473686?utm_source=mdpi.com&utm_medium=website&utm_campaign=avatar_name)
- [Recommended Articles](https://www.mdpi.com/1999-4907/17/4/419)
- [Related Info Link](https://www.mdpi.com/1999-4907/17/4/419#related)
- [Google Scholar](http://scholar.google.com/scholar?q=Land%20Surface%20Phenology%20Reveals%20Region-Specific%20Hurricane%20Impacts%20Across%20the%20North%20Atlantic%20Basin%20%282001%E2%80%932022%29)
- [More by Authors Links](https://www.mdpi.com/1999-4907/17/4/419#authors)
- [on DOAJ]()
- [Topete-Pozas, C.](http://doaj.org/search/articles?source=%7B%22query%22%3A%7B%22query_string%22%3A%7B%22query%22%3A%22%5C%22Carlos%20Topete-Pozas%5C%22%22%2C%22default_operator%22%3A%22AND%22%2C%22default_field%22%3A%22bibjson.author.name%22%7D%7D%7D)
- [Norman, S. P.](http://doaj.org/search/articles?source=%7B%22query%22%3A%7B%22query_string%22%3A%7B%22query%22%3A%22%5C%22Steven%20P.%20Norman%5C%22%22%2C%22default_operator%22%3A%22AND%22%2C%22default_field%22%3A%22bibjson.author.name%22%7D%7D%7D)
- [on Google Scholar]()
- [Topete-Pozas, C.](http://scholar.google.com/scholar?q=Carlos%20Topete-Pozas)
- [Norman, S. P.](http://scholar.google.com/scholar?q=Steven%20P.%20Norman)
- [on PubMed]()
- [Topete-Pozas, C.](http://www.pubmed.gov/?cmd=Search&term=Carlos%20Topete-Pozas)
- [Norman, S. P.](http://www.pubmed.gov/?cmd=Search&term=Steven%20P.%20Norman)
[Article Views](https://www.mdpi.com/1999-4907/17/4/419#metrics)
[Citations \-](https://www.mdpi.com/1999-4907/17/4/419#metrics)
- [Table of Contents](https://www.mdpi.com/1999-4907/17/4/419#table_of_contents)
Altmetric [*share* Share](https://www.mdpi.com/1999-4907/17/4/419 "Share") [*announcement* Help](https://www.mdpi.com/1999-4907/17/4/419 "Help") [*format\_quote* Cite]() [*question\_answer* Discuss in SciProfiles](https://sciprofiles.com/discussion-groups/public/10.3390/f17040419?utm_source=mpdi.com&utm_medium=publication&utm_campaign=discuss_in_sciprofiles "Discuss in Sciprofiles")
## Need Help?
### Support
Find support for a specific problem in the support section of our website.
[Get Support](https://www.mdpi.com/about/contactform)
### Feedback
Please let us know what you think of our products and services.
[Give Feedback](https://www.mdpi.com/feedback/send)
### Information
Visit our dedicated information section to learn more about MDPI.
[Get Information](https://www.mdpi.com/authors)
[*clear*]()
## JSmol Viewer
[*clear*]()
*first\_page*
[Download PDF](https://www.mdpi.com/1999-4907/17/4/419/pdf?version=1774604120)
*settings*
[Order Article Reprints](https://www.mdpi.com/1999-4907/17/4/419/reprints)
Font Type:
*Arial* *Georgia* *Verdana*
Font Size:
Aa Aa Aa
Line Spacing:
** ** **
Column Width:
** ** **
Background:
Open AccessFeature PaperArticle
# Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022)
by
Carlos Topete-Pozas
Carlos Topete-Pozas
[SciProfiles](https://sciprofiles.com/profile/3594599?utm_source=mdpi.com&utm_medium=website&utm_campaign=avatar_name) [Scilit](https://scilit.com/scholars?q=Carlos%20Topete-Pozas) [Preprints.org](https://www.preprints.org/search?condition_blocks=[{%22value%22:%22Carlos+Topete-Pozas%22,%22type%22:%22author%22,%22operator%22:null}]&sort_field=relevance&sort_dir=desc&page=1&exact_match=true) [Google Scholar](https://scholar.google.com/scholar?q=Carlos+Topete-Pozas)
[](https://orcid.org/0000-0003-0240-2823) and
Steven P. Norman
Steven P. Norman
[SciProfiles](https://sciprofiles.com/profile/3441195?utm_source=mdpi.com&utm_medium=website&utm_campaign=avatar_name) [Scilit](https://scilit.com/scholars?q=Steven%20P.%20Norman) [Preprints.org](https://www.preprints.org/search?condition_blocks=[{%22value%22:%22Steven+P.+Norman%22,%22type%22:%22author%22,%22operator%22:null}]&sort_field=relevance&sort_dir=desc&page=1&exact_match=true) [Google Scholar](https://scholar.google.com/scholar?q=Steven+P.+Norman)
\*[](https://orcid.org/0000-0003-2080-9774)
US Department of Agriculture Forest Service, Southern Research Station, Eastern Forest Environmental Threat Assessment Center, 200 WT Weaver Blvd, Asheville, NC 28804, USA
\*
Author to whom correspondence should be addressed.
*Forests* **2026**, *17*(4), 419; <https://doi.org/10.3390/f17040419>
Submission received: 24 February 2026 / Revised: 17 March 2026 / Accepted: 25 March 2026 / Published: 27 March 2026
(This article belongs to the Special Issue [Innovating Indicators: New Approaches for Tracking Forest Health Status and Trends in a Rapidly Changing World](https://www.mdpi.com/1999-4907/17/4/%20%20%20%20%20%20%20%20%0A%20%20%20%20/journal/forests/special_issues/OB6XFCL62N%0A))
[Download *keyboard\_arrow\_down*]()
[Download PDF](https://www.mdpi.com/1999-4907/17/4/419/pdf?version=1774604120)
[Download PDF with Cover](https://www.mdpi.com/1999-4907/17/4/419)
[Download XML](https://www.mdpi.com/1999-4907/17/4/419)
[Download Epub](https://www.mdpi.com/1999-4907/17/4/419/epub)
[Browse Figures](https://www.mdpi.com/1999-4907/17/4/419)
[Versions Notes](https://www.mdpi.com/1999-4907/17/4/419/notes)
## Abstract
Hurricanes routinely damage forests across the North Atlantic Basin, yet efforts to characterize their impacts have had mixed subregional success. To elucidate these challenges, this study analyzed pre- and post-hurricane land surface phenology (LSP) for 44 moderate and strong hurricanes over 22 years using the Enhanced Vegetation Index (EVI). We statistically grouped storms based on their long-term climate attributes, then compared subregional impacts with wind speed and land cover. After accounting for wind speed, responses differed among the six subregions. The Southeast U.S. showed declines in EVI for the first winter and first year post storm, but this response was weak or absent elsewhere. The Central America region declined in the first winter but not in the subsequent growing season, while four other regions showed no increased impact with wind speed in either season. We then examined six category 4 hurricanes using a forest mask. In dry areas, drought-sensitive vegetation explained weak responses, whereas in the humid tropics, rapid refoliation or sprouting was common. These factors complicate optical remote sensing assessments. Rapid evaluations can mistake defoliation for more substantial damage, and delayed assessments can confuse EVI recovery with structural recovery. Results underscore the need for ecologically tailored monitoring approaches.
Keywords:
[hurricanes](https://www.mdpi.com/search?q=hurricanes); [land surface phenology](https://www.mdpi.com/search?q=land+surface+phenology); [EVI](https://www.mdpi.com/search?q=EVI); [land cover](https://www.mdpi.com/search?q=land+cover); [climate variability](https://www.mdpi.com/search?q=climate+variability)
## 1\. Introduction
The North Atlantic Basin (NAB) commonly experiences extreme hurricanes with severe consequences for the natural and human systems of the region \[[1](https://www.mdpi.com/1999-4907/17/4/419#B1-forests-17-00419),[2](https://www.mdpi.com/1999-4907/17/4/419#B2-forests-17-00419),[3](https://www.mdpi.com/1999-4907/17/4/419#B3-forests-17-00419),[4](https://www.mdpi.com/1999-4907/17/4/419#B4-forests-17-00419)\]. Notably, the intensity and frequency of hurricanes in this region have been changing over the recent decades \[[3](https://www.mdpi.com/1999-4907/17/4/419#B3-forests-17-00419),[5](https://www.mdpi.com/1999-4907/17/4/419#B5-forests-17-00419),[6](https://www.mdpi.com/1999-4907/17/4/419#B6-forests-17-00419)\]. Between 1989 and 2018, an estimated one million km2 of forest and other land cover types were affected by hurricanes in the NAB, representing one of the highest damaged areas globally \[[7](https://www.mdpi.com/1999-4907/17/4/419#B7-forests-17-00419)\]. Indeed, the majority of studies regarding hurricane impacts on forest (67%) were conducted in this region \[[7](https://www.mdpi.com/1999-4907/17/4/419#B7-forests-17-00419)\].
Remote sensing has become a critical tool for assessing hurricane impacts and subsequent vegetation recovery at different spatio-temporal scales \[[8](https://www.mdpi.com/1999-4907/17/4/419#B8-forests-17-00419),[9](https://www.mdpi.com/1999-4907/17/4/419#B9-forests-17-00419)\]. Its usefulness for monitoring within the NAB is influenced by complex patterns of land cover and seasonal vegetation dynamics, both of which are informed by land surface phenology (LSP). Expanded use of LSP has improved how we understand disturbance impacts in light of seasonal and multi-year environmental and anthropogenic influences \[[10](https://www.mdpi.com/1999-4907/17/4/419#B10-forests-17-00419),[11](https://www.mdpi.com/1999-4907/17/4/419#B11-forests-17-00419),[12](https://www.mdpi.com/1999-4907/17/4/419#B12-forests-17-00419)\]. LSP-derived measures relate to important ecosystem processes, such as carbon and water cycles, timber markets, the fuels available for wildfire, and species interactions and connectivity \[[13](https://www.mdpi.com/1999-4907/17/4/419#B13-forests-17-00419)\]. LSP is particularly useful for tracking vegetation productivity, and this varies regionally and locally with vegetational composition, climate, soil, management practices, and weather disturbances, among other factors \[[14](https://www.mdpi.com/1999-4907/17/4/419#B14-forests-17-00419)\].
Recent research demonstrates the broad monitoring capabilities of LSP to characterize damage from hurricanes. Continuous satellite measurements give researchers the ability to track commonly used LSP parameters that capture key stages of seasonal vegetation behavior, such as the start of spring or fall or the peak of the growing season. In tropical or subtropical climates, hurricanes can impact the onset or progression of the late growing season contingent on when storms occur. This limitation makes LSP parameters that capture seasonal timing less reliable. Instead, for consistency, post-storm season integral phenometrics are more suitable, such as area under the curve (AUC) for all or a portion of the year. Such measures are recognized LSP phenometrics that provide a robust mechanism for understanding seasonal and annual variability \[[10](https://www.mdpi.com/1999-4907/17/4/419#B10-forests-17-00419),[15](https://www.mdpi.com/1999-4907/17/4/419#B15-forests-17-00419),[16](https://www.mdpi.com/1999-4907/17/4/419#B16-forests-17-00419),[17](https://www.mdpi.com/1999-4907/17/4/419#B17-forests-17-00419)\]. Along the US coast, studies have demonstrated the value of short- and long-term monitoring using different sensors and spectral indices including the 10 m resolution Sentinel 2 Normalized Difference Vegetation Index (NDVI) \[[18](https://www.mdpi.com/1999-4907/17/4/419#B18-forests-17-00419)\], 30 m resolution Landsat 5 and 7 and NDVI \[[19](https://www.mdpi.com/1999-4907/17/4/419#B19-forests-17-00419)\], and five vegetation indices from MODIS 1 km 16-day \[[20](https://www.mdpi.com/1999-4907/17/4/419#B20-forests-17-00419)\]. Others \[[21](https://www.mdpi.com/1999-4907/17/4/419#B21-forests-17-00419)\] assessed the effects of Hurricane Maria in Puerto Rico using Landsat 8 data with spectral mixture and statistical analyses in association with landform characteristics and forest structure.
Enhanced Vegetation Index (EVI) has been considered by several authors as a more suitable index to monitor forest disturbance in tropical regions because of EVI’s sensitivity to canopy variation in high biomass regions where NDVI saturates \[[22](https://www.mdpi.com/1999-4907/17/4/419#B22-forests-17-00419),[23](https://www.mdpi.com/1999-4907/17/4/419#B23-forests-17-00419),[24](https://www.mdpi.com/1999-4907/17/4/419#B24-forests-17-00419)\]. Research has demonstrated the effectiveness of MODIS EVI to assess the extent of immediate impacts from Hurricane Dean in the tropical forest of the Yucatan Peninsula, Mexico \[[25](https://www.mdpi.com/1999-4907/17/4/419#B25-forests-17-00419)\], and in Central America after Hurricane Felix \[[26](https://www.mdpi.com/1999-4907/17/4/419#B26-forests-17-00419)\]. Others \[[27](https://www.mdpi.com/1999-4907/17/4/419#B27-forests-17-00419)\] used MODIS EVI to assess hurricane damage and recovery in the northern Gulf and found that impacts vary by event, year, vegetation characteristics, and climate attributes \[[28](https://www.mdpi.com/1999-4907/17/4/419#B28-forests-17-00419)\], and monitored forest recovery along the northern Gulf comparing MODIS EVI with other spectral indices. In the Caribbean, ref. \[[29](https://www.mdpi.com/1999-4907/17/4/419#B29-forests-17-00419)\] used a MODIS disturbance index to capture both drought and hurricane impacts over four large Caribbean islands, and stressed the importance of considering land cover. In Cuba, ref. \[[30](https://www.mdpi.com/1999-4907/17/4/419#B30-forests-17-00419)\] assessed the extent and severity of damage to mangroves after Hurricane Irma with integrated use of MODIS and Sentinel-2 data. MODIS has been useful for addressing hurricane impacts in the tropics because its high temporal frequency helps overcome issues with cloud cover while isolating immediate responses and longer-term implications.
The NAB’s complex meteorological and land cover characteristics present challenges for monitoring hurricane impacts as methods may not be uniformly effective \[[31](https://www.mdpi.com/1999-4907/17/4/419#B31-forests-17-00419),[32](https://www.mdpi.com/1999-4907/17/4/419#B32-forests-17-00419)\]. While rapid optical remote sensing assessments are broadly effective for recognizing the footprint of potential impacts, commonly used techniques struggle with ecological precision, such as lasting structural damage and successional impacts in the tropics. Assessments in the humid tropics are especially time-sensitive with a high risk of misinterpreting recovery of the remote sensing index as something more. The most intensive research has taken place in the southern U.S., and this regional bias limits our understanding of monitoring effectiveness across the broader region \[[7](https://www.mdpi.com/1999-4907/17/4/419#B7-forests-17-00419)\]. The broader NAB region includes areas with substantially different vegetation and climate variability that may require different approaches for hurricane impact assessments to be accurate. Our research addresses these regional differences in the effectiveness of LSP monitoring after hurricanes to clarify important limitations and research needs. Our specific objectives include evaluating the importance of wind speed, land cover, and climate through the use of regional assessment and case studies.
## 2\. Materials and Methods
### 2\.1. Study Area and Hurricanes
We focused on major storms rather than tropical storms and Saffir–Simpson category 1 hurricanes to increase the likelihood of capturing wind damage to forests. We selected 44 category 2–5 hurricanes that made landfall within the NAB between the years 2001 to 2022 across the Southeast U.S., Mexico, the Caribbean, and Central America. We determined the landfall site and inland path for each hurricane using the International Best Track Archive for Climate Stewardship (IBTrACS) dataset \[[33](https://www.mdpi.com/1999-4907/17/4/419#B33-forests-17-00419)\]. For each hurricane, we defined a semi-uniform focal area by drawing a polygon that extended 25 km on each side of the track and 100 km inland. This standardized footprint is intended to emphasize the near-track region most likely to experience inner-core winds and associated severe wind damages and provides a consistent spatial basis for comparable impact estimation across storms and regions. In some cases, such as in the Caribbean, the polygons were somewhat smaller due to the size of the island or the need to exclude water bodies. We used each footprint to derive MODIS EVI data, land cover, and climate attributes. [Figure 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f001) shows an overview of the data used and workflow.
### 2\.2. Climate Regionalization
Various researchers have classified the NAB-related tropical cyclones in subregions according to ecology or climate factors \[[34](https://www.mdpi.com/1999-4907/17/4/419#B34-forests-17-00419),[35](https://www.mdpi.com/1999-4907/17/4/419#B35-forests-17-00419)\]. Our need was to characterize the basic climate drivers that influenced vegetation and recovery response, so instead of relying on these subdivisions, we derived gridded climate variables for each focal area directly. For each hurricane focal area, we derived the median monthly precipitation, maximum temperature, and vapor pressure deficit from 2000 to 2022 from the Climate Engine TerraClimate 4 km-Monthly dataset \[[36](https://www.mdpi.com/1999-4907/17/4/419#B36-forests-17-00419)\]. We then combined these climate time series for each focal area and analyzed their cross-seasonal and interannual similarities using non-hierarchical k\-means clustering \[[37](https://www.mdpi.com/1999-4907/17/4/419#B37-forests-17-00419)\] using R 2024.04.2 software \[[38](https://www.mdpi.com/1999-4907/17/4/419#B38-forests-17-00419)\]. We explored prescribing different numbers of clusters to balance the need for representation with geographic coherence. The final iteration assigned a statistically-derived membership for each of our 44 hurricanes to six distinct climate subregions.
### 2\.3. Hurricane Impacts and Their Drivers
High-frequency, multi-year LSP data broadly captures recurring seasonal behavior, hurricane impacts, and recovery. For each of the 44 focal areas, we calculated the areal mean 16-day EVI for a 22 year period (2001–2022) using 500 m MODIS Terra (MOD13A1) and Aqua (MYD13A1) data derived from Climate Engine (visited on July 2023). Instances of missing data from the source were handled using a non-parametric imputation method implemented in the missForest package in R 2024.04.2 software \[[38](https://www.mdpi.com/1999-4907/17/4/419#B38-forests-17-00419),[39](https://www.mdpi.com/1999-4907/17/4/419#B39-forests-17-00419)\]. This algorithm employs Random Forests to estimate and replace missing values in a dataset. To document the normal seasonal LSP behavior for each cluster across years, we combined the 22 years of 16-day EVI values for the member hurricanes for an annual curve showing the mean and standard deviation. We then compared the dates of the hurricanes for each cluster with their annual LSP to clarify the implications of LSP seasonality for monitoring.
To understand impacts for each focal area, we used the MODIS Terra & Aqua 500 m 16-day dataset. We clipped the EVI time series to have two full years before and two years after each storm’s calendar year and derived two seasonal measures: the area under the curve (AUC) for the first winter (December through February) and the AUC for the first full year after the storm. Impacts were quantified as the percent change in EVI AUC (dEVI) from the mean of the two pre-storm seasonal values. For each subregional cluster, we used regression to compare measured impacts to the respective hurricane’s median wind speed that was derived from IBTrACS \[[33](https://www.mdpi.com/1999-4907/17/4/419#B33-forests-17-00419)\]. We then performed Mann–Whitney (Wilcoxon) tests on the regressions to compare differences among regions for the two seasonal measures.
Land cover varied among the focal areas and this could potentially influence the vulnerability of vegetation to the storms and our interpretation of short and long-term impacts. These patterns and differences among focal areas were assessed in Google Earth Engine using the MCD12Q1.061 MODIS Land Cover Type Yearly Global 500 m dataset. We summarized the percent land cover for each subregional cluster. To simplify categories across the region we reclassified the original classes into seven broader groups: mixed forest and woody savannas were merged into “mixed forest”; urban, water, wetland and other were grouped as “other”; and croplands, savannas, evergreen broadleaf forest and deciduous broadleaf forest were used as defined in the original dataset.
### 2\.4. Local Case Studies
To achieve a more precise understanding of impacts for the six subregions, we analyzed the LSP on one representative category 4 hurricane for each subregional cluster. Hurricane Charley made landfall on 13 August 2004 near Cayo Costa, Florida, with a sustained wind of 210–249 km/h and caused an estimated \$14 billion in economic losses. Hurricane Laura made landfall near Cameron, Louisiana on 27 August 2020 causing an estimated of \$19 billion in economic losses. Hurricane Harvey, that was considered the second costliest U.S. tropical storm after Katrina, struck Texas on 26 August 2017 and caused an estimated damage of 125 billion. Hurricane Iota made landfall in Nicaragua on 17 November 2020, just two weeks after Eta passed over this same area. Hurricane Emily made landfall in the Yucatan Peninsula with sustained winds of 212 km/h. Hurricane Maria struck Puerto Rico on 20 September 2017 and resulted in severe damage. Detailed information on hurricane history and damage costs can be found at (<https://www.weather.gov/publications/assessments>, accessed on 24 February 2026, <https://www.noaa.gov/>, accessed on 24 February 2026).
For each of these six hurricanes, we derived a median EVI time series using the MODIS Terra (MOD13Q1) and Aqua (MYD13Q1) 250 m 16-day dataset and, respectively, that was compiled into an 8-day time series for all lands and forests alone. Our forest cover mask used the Global Land Cover ESA WorldCover 10 m v100 map (2020) dataset and was coarsened to 250 m. We smoothed each time series using a Savitsky–Golay algorithm in R 2024.04.2 software using a window of 23 observations to reduce noise \[[38](https://www.mdpi.com/1999-4907/17/4/419#B38-forests-17-00419),[40](https://www.mdpi.com/1999-4907/17/4/419#B40-forests-17-00419),[41](https://www.mdpi.com/1999-4907/17/4/419#B41-forests-17-00419),[42](https://www.mdpi.com/1999-4907/17/4/419#B42-forests-17-00419)\]. Masking allowed us to resolve the potential influence of land cover on LSP for these six case studies and to help interpret results from the prior analysis of the 44 hurricanes. We calculated similar measures of impact as we had for the prior analysis. Percent change in EVI AUC was calculated as the difference between the mean of the two years prior to the hurricane compared to the year after the event. The EVI AUC of the first winter after the storm (December to February) was compared the mean of the prior two winters to minimize the effects of interannual variability.
## 3\. Results
### 3\.1. Climate Regionalization
Clustering grouped the 44 hurricanes into six types defined by climate, which we labeled according to their primary subregional location ([Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002), [Table 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t001)). The Southeast U.S. (SUS) type had the most hurricanes at 13 and with all categories and a variety of wind speeds ranging from 148 to 259 km/h. The Caribbean (C) group included 10 hurricanes ranging from category 2 to 4. The Florida Peninsula (FP), South Gulf (SG) and Central America (CA), had six hurricanes each. The FP showed hurricanes categories 2 to 4 and the latter two from 2 to 5. With only three hurricanes represented, the West Gulf group had the fewest hurricanes, and these ranged from category 2 to 4. Clustering yielded strong spatial aggregations for FP, SUS, WG and CA. The location of C and SG were less well defined spatially, with the C group extending from Puerto Rico and Cuba to Belize and Mexico.
Land cover differed among the six subregions with respect to woody (deciduous, mixed, and evergreen forests) and other types ([Figure 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f003)). SG and CA averaged the most forest, with SUS and C averaging around 50% forest and FP and WG being predominantly savanna and grassland.
Subregions with the most forested land cover (SG and CA) show the highest median EVI while that with the least woody vegetation (WG) has the lowest EVI ([Figure 4](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f004)a). These differences in EVI are also consistent with the climate attributes of the subregions. CA’s high precipitation, warm temperature, and low VPD reflect its low latitude climate ([Figure 4](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f004)b–d). This contrasts with the least productive subregion, WG, that has the lowest precipitation and highest VPD ([Figure 4](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f004)b,d). SUS shows a broad range of EVI values and strong variation in temperature that reflects the strong seasonality of the higher-latitude climate and the large number of hurricanes included.
For all six subregions, hurricanes occurred near or following the peak of the growing season when vegetation was in normal seasonal decline ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)). The median and standard deviation for each subregion shows the interannual variation over the 21-year period for the member hurricanes. Tropical CA had the highest year-round median EVI that consistently exceeded 0.5. SG had considerably more seasonal variability despite peaking with a higher EVI in July at the onset of its hurricane season. C shows a low median EVI below 0.4 early in the year, with the highest values around 0.55 during summer, indicating relatively moderate productivity with pronounced seasonality. Of all the subregions, C has the broadest range of hurricane dates represented, which is consistent with the Caribbean’s geographic exposure. FP and SUS show lower peak EVIs with strong seasonal variation, and WG exhibits the lowest median EVI values of all six subregions with a weak growing season peak that is comparatively stable between May and October when the three hurricanes occurred.
### 3\.2. Hurricane Impacts
Of the six subregions, only SUS, FP and WG showed declines in EVI for both the first winter and first growing season after the storm, and the decline increased across those two periods ([Table 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t002)). While higher than expected for winter, C also shows a potential hurricane-associated decline for the first year after the storm. In contrast, more tropical CA showed the strongest mean decline of any subregion for the first winter, but none for the subsequent growing season. SG showed no measured decline on average, as EVI increased for both seasonal measures. Of all the subregions, SUS’s decline was most consistent across seasons.
Wind speed and EVI decline show an inconsistent relationship across subregions. For the first winter after the storms, stronger winds were associated with more decline for SUS (p = 0) and CA (p = 0.005), but other subregions showed no anticipated decline or pattern ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)a). For the first year after the storm, SUS consistently showed a decline in impacts with increased wind speed (p = 0), but no other subregion showed this relationship ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)b).
### 3\.3. Local Case Studies
The six selected category 4 hurricanes show generally consistent results for all lands compared to their respective subregions ([Table 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t002) and [Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003)). Five of the six hurricanes declined for the first winter, with Emily (SG) being the exception. Hurricane Maria (C) differed from its subregion’s response by also showing a first winter decline. For the first year after the storm, neither Maria (C) nor Charley (FP) showed a decline, but Iota (CA) did, although much less than in winter and unlike the response of the CA subregion overall. From the first winter to the first year, half of the cases show further decline while half improved, with Iota (CA) showing the most remarkable improvement.
The isolated forest responses of these six cases are directionally similar to their corresponding all-land responses but often with very different magnitudes ([Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003)). All cases but Emily (SG) showed an initial drop in forest EVI, which was a directionally consistent but stronger increase than its all-lands response. Note that Emily occurred in mid-July, which was one to four months earlier than the dates of the other five cases ([Table 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t001)), and has extensive broadleaf forest cover ([Figure 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f003)). Iota (CA)’s forest decline was much stronger than its all-lands decline, but its EVI first-year recovery was greater. Maria (C) and Charley (FP) appear to have also recovered quickly. The forest EVI behavior of Iota, Maria and Charley is the opposite of what occurred with Laura (SUS), Harvey (WG) and Emily (SG); each of these showed increased EVI departure from the first winter to the first year after the storm.
Comparison of the five-year LSP behavior for forests and all lands shows a broadly similar form prior to the storms with Iota (CA) being most different ([Figure 7](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f007)). Hurricane impacts are less than the seasonal variation in EVI. There are often recurring differences in cover types in terms of their seasonal amplitude or peak timing. Note in particular the earlier forest EVI peak for Laura (SUS), Harvey (WG), and Emily (SG).
As LSP measures can vary among years for reasons other than disturbance such as seasonal weather, the EVI profiles for the two pre-hurricane years shown in [Figure 7](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f007) can inform the interpretation and attribution of change shown in [Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003). Note the more irregular two-year pre-storm LSP variability for Harvey (WG), Iota (CA), Emily (SG), and Maria (C) for the first winter or full year. Also note the more complex, less predictable LSP behavior of both forest and all lands for Iota (CA) over all five years.
## 4\. Discussion
In this study, we investigated the impacts of 44 hurricanes that made landfall across six distinct subregions of the NAB using remote sensing. We show that seasonal insights from LSP help clarify and nuance our understanding of hurricane impacts, given the major climatic and ecological differences that are present across the region. Using two measures, the first winter after the storm and the first year after the storm, we demonstrated that higher wind speeds do not consistently correlate with greater storm impacts. For example, significant declines in EVI with increasing wind speed were observed only in the Southeast U.S., while other subregions showed weak or non-significant relationships ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)). Phenologically, when hurricanes occurred during the mid growing season, as indicated on [Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005), refoliation and rapid EVI recovery may have reduced our ability to recognize the magnitude of structural impacts. These ecological and technological limitations suggest the quality of remote sensing assessments using this common technique, particularly for tropical areas that have had the least field validations.
Isolating land cover types is a necessity. Non-woody vegetation types, such as grasslands or pasture, are biologically insensitive to wind damage, but they are often hypersensitive to unrelated climate variation that drives annual productivity. For analyses of change over time, the presence of these climate-sensitive types can distort the reliability of baseline conditions and the integrity of the change measure used to evaluate the post-storm vegetation state. This is a hazard to mixed land cover assessments when coarse-grid imagery is used, as non-targeted vegetation behavior can influence results. In heterogeneous landscapes, forest impacts may only be isolated through use of higher resolution imagery and masking that targets the susceptible vegetation. However, even when a high-resolution forest mask is available, it can be challenging to obtain high-resolution cloud-free imagery to conduct assessments at scale.
The phenological status of a forest at the time of the storm can affect the accuracy of impact assessments. An immediate response may indicate either structural tree damage or ephemeral leaf stripping, given the sensitivity of most spectral indices to foliage. Seasonally deciduous forest types may be particularly vulnerable to leaf stripping late in their growing season when their foliage is fragile due to senescence, and that is when most hurricanes occur ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)). If the goal is to map structural damage in a rapid assessment, not just the general footprint of the storm more generally, leaf stripping can give false positives—a hazard that can be further aggravated with change detection approaches when the timing of senescence varies from year to year. To avoid this, we conducted our first impact assessment during the first winter rather than immediately after the storm ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)), but this may have introduced a different problem as storms that occurred during the mid growing season may have experienced rapid refoliation prior to the winter assessment that persists. In the tropics, having a different time-since-storm could have affected the consistency of our first-winter measure among hurricanes. Remeasuring impacts across more than the two seasons used in this study may provide a way to isolate these complex phenological factors, but that can be difficult to achieve at scale in areas with frequent cloud cover. Very rapid assessments can also be misleading when hurricane rain provides drought relief through a short-lived resurgence of stressed vegetation, and this is a particularly confounding factor for areas with mixed land cover, pasture and grass. It is unclear if this rain response affected our results given the time of our winter assessment, but it could help explain surges of greenness reported by others using NDVI time series \[[43](https://www.mdpi.com/1999-4907/17/4/419#B43-forests-17-00419)\].
Crown and bole sprouting can initiate just weeks after a hurricane in tropical forests, and this phenomenon can obscure structural damage \[[44](https://www.mdpi.com/1999-4907/17/4/419#B44-forests-17-00419),[45](https://www.mdpi.com/1999-4907/17/4/419#B45-forests-17-00419)\]. This phenomenon is less relevant for portions of SUS where hurricane damage to the predominant industrial loblolly pine forests often leads to tree mortality \[[46](https://www.mdpi.com/1999-4907/17/4/419#B46-forests-17-00419)\]. However, even north of SUS in the Appalachians, hardwood sprouting was prevalent after 2024’s hurricane Helene during the first growing season after the storm (personal observations of the second author, July 2025), suggesting that this is a widespread wind adaptive response. Sprouting vegetation can reduce the accuracy of impact assessments that use optical remote sensing, and this limits our understanding of forest resistance and recovery. Used alone or in combination, alternative technologies that capture change in vegetation height, such as high-resolution aerial LiDAR, largely bypass much of this problem as sprouting can occur on crown-damaged, leaning and fallen trees. Unfortunately, precise structural assessments using this technology take time and the extent of the analysis can be restricted \[[47](https://www.mdpi.com/1999-4907/17/4/419#B47-forests-17-00419)\].
### Subregions
Our six subregions differ in terms of climate variability, land cover and vegetational sensitivity, and phenology. These attributes are ecologically inter-related and they provide useful ways to frame post-hurricane impact assessments. Awareness of these subregional attributes can help identify where research is lacking and help tailor post-hurricane monitoring to improve its effectiveness.
The Southeastern U.S. (SUS) stands out as a reliable area for long- and short-term monitoring of hurricane effects, and this is consistent with prior studies \[[7](https://www.mdpi.com/1999-4907/17/4/419#B7-forests-17-00419),[48](https://www.mdpi.com/1999-4907/17/4/419#B48-forests-17-00419),[49](https://www.mdpi.com/1999-4907/17/4/419#B49-forests-17-00419)\]. Impacts are resolvable during both winter and the subsequent growing season with impacts increasing at higher wind speeds, as expected ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)). Other research has shown that damage to the industrial pine of this subregion is especially well-captured, as is the post-hurricane management response, but impacts to deciduous and mixed forests are harder to resolve \[[18](https://www.mdpi.com/1999-4907/17/4/419#B18-forests-17-00419)\]. Isolating leaf stripping from structural damage for deciduous forests can be challenging for storms that occur during fall senescence, but strong winter seasonality limits the potential for rapid hardwood sprouting.
In contrast, monitoring hurricane impacts in Central America (CA) is more challenging, particularly for the year after the storm. This is apparently due to the warm-wet tropical climate and high ecosystem productivity that causes rapid EVI recovery for the evergreen broadleaf and mixed forest types. For the first winter after the storm, we documented EVI declines that strengthened with wind speed, but we found no consistent evidence of impacts thereafter. If extreme hurricanes cause lasting structural damage to these forests, it is likely being obscured by refoliation, because over 70% of the land cover of CA is forest. Category 4 Iota had the strongest initial departure of any of the 44 hurricanes considered, but despite category 5 winds, Felix had less impact, although its impact was directionally similar to Iota over the two periods ([Table 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t001), [Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)). This may be because Felix’s focal area had substantially more grassland which is less sensitive to damage than forest ([Appendix A](https://www.mdpi.com/1999-4907/17/4/419#app1-forests-17-00419), [Table A1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t0A1)). The irregular year-to-year LSP shown for Hurricane Iota ([Figure 7](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f007)d) complicates assessment, and this may be caused by the quality of the remote sensing data given the frequency of clouds in this humid region. In CA, hurricanes tend to occur in late November, which is phenologically later than that other subregions ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)). That difference in timing may affect the sensitivity of the deciduous canopy to defoliation or how aggressively hardwood refoliation occurs. Monitoring damage during first winter may be the most feasible for the CA subregion, but it is challenging to isolate structural impacts.
Only three of 44 hurricanes were included in the West Gulf (WG) subregion, and Harvey, the strongest, had the lowest mean wind speed of the six category 4 case studies ([Table 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t001)). Consistent with its warm-dry climate, WG had the lowest EVI across seasons and the least forest cover ([Figure 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f003), [Figure 4](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f004) and [Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005), [Appendix A](https://www.mdpi.com/1999-4907/17/4/419#app1-forests-17-00419), [Table A1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t0A1)). At less than 10% forest, the subregion’s vegetation should be inherently less sensitive to hurricane damage, given that grass is unlikely to retain damage into the next growing season, yet both forests and all lands show notable declines ([Table 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t002) and [Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003)). It is particularly surprising that the first-year departure of all land cover is roughly double that of the forest for Harvey ([Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003)). This suggests a cause from something other than the storm, such as the influence of interannual climate variability. Grass is particularly sensitive to climate variability and drought \[[50](https://www.mdpi.com/1999-4907/17/4/419#B50-forests-17-00419),[51](https://www.mdpi.com/1999-4907/17/4/419#B51-forests-17-00419)\]. Our measure of storm impact, namely, change from a prior baseline, is particularly vulnerable to seasonal climate variability because either the pre- or post-disturbance value could be compromised. This problem appears to explain the first-year response of 2010’s category 2 Alex that, contrary to expectations, showed a stronger decline than the more intense storms for this subregion. Unlike the baseline used to calculate change, 2011’s first-year EVI was likely low given a severe drought as indicated by the North American Drought Monitor (<https://droughtmonitor.unl.edu/NADM/Maps.aspx>, accessed on 24 February 2026). This inherent limitation of monitoring in the subregion can reduce confidence in the accuracy of impact assessments, including those that rely on coarse-resolution forest masks. Results are also more challenging to interpret in this subregion with so few storms represented. For this subregion, forest monitoring requires a high spatial resolution analysis and sensitivity to interannual climate variability.
The South Gulf (SG) subregion is characterized by highly productive, seasonal forests with more than 80% forest cover, the highest of the six subregions ([Figure 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f003)). Annual EVI is the second-highest among the subregions, peaking as the highest but with a strong winter decline that shows high variability ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)). Hurricanes in SG occurred from July, at peak EVI, through September, which is phenologically earlier than for those of other subregions. Rapid refoliation occurs during the growing season, and early storm dates may reduce the usefulness of first winter and first year measures ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)). For example, Emily, the subregion’s case study, was the earliest of the six cases, and the fourth-earliest of all 44 storms analyzed ([Table 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t001)). This adaptation of tropical hardwoods may be more strongly expressed when defoliation occurs during the mid growing season, so impact assessment may be more accurate when conducted immediately after the hurricane occurs, although monitoring at this time can be difficult for low lying areas due to lingering floodwaters and the prevalence of non-structural defoliation.
The ten hurricanes analyzed for the Caribbean (C) subregion show varied responses for the first winter and first year after their storms with respect to wind speed. Some storms show a decline while others show an increase in EVI for these two periods. A recent study from Turner et al., \[[30](https://www.mdpi.com/1999-4907/17/4/419#B30-forests-17-00419)\], on the effects of hurricane Irma in Cuba found that while there was a significant decrease in EVI in mangroves and wetlands, there was a widespread increase in EVI elsewhere, including for the dry forest. In our analysis, Maria showed an EVI decline for the first winter after the storm, but an increase in EVI during the first year after the storm. This behavior is consistent with de Beurs et al., \[[29](https://www.mdpi.com/1999-4907/17/4/419#B29-forests-17-00419)\] who documented decline from Maria in Puerto Rico, with a quick recovery starting 8–12 weeks after the storm which they suggested might be associated with a growth surge from heavy rainfall. The timing of hurricanes here with respect to the region’s phenology suggests a strong potential for rapid refoliation and growth before the end of the growing season much of the time ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)). In addition, with less forest cover than SG and CA and rugged topography that fragments forest cover, EVI’s sensitivity to post-storm weather variability may also be high. Similarly to SG, resolving hurricane impacts in this subregion can be challenging, but insights are possible soon after the storm. Use of high resolution imagery can help isolate the confounding responses of cover types.
Only about 10 percent of the Florida Peninsula (FP) subregion is forested, as savannas, grasslands, croplands, and other land cover types predominate. These latter types are less sensitive to damage, and this likely explains the non-responsiveness of EVI regardless of wind speed ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)). The response of FP’s Hurricane Charley shows a short-term decline followed by rapid recovery for both all lands and forests, and this behavior is generally consistent with our detailed case studies for C and CA ([Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003)). This may reflect the rapid sprouting of trees and shrubs that may or may not have substantial structural damage. While limited in distribution, hurricane damage to south Florida’s mangrove forests has received the most research attention and the homogeneity of the type makes it more easily monitored. The mangroves here have had four declines from severe hurricanes during the last three decades with slow recovery and a long-term decline \[[52](https://www.mdpi.com/1999-4907/17/4/419#B52-forests-17-00419)\]. Elsewhere, successful hurricane monitoring in the FP will likely benefit from high-resolution analyses that isolate the patchy woody plant cover.
## 5\. Conclusions
This study identified inconsistencies in the predicted relationship between wind speed and hurricane damage. Subtropical SUS showed increased damage with wind speed—a pattern that remained across seasons. This anticipated relationship was absent elsewhere, as EVI change often increased after severe disturbance. Confounding drivers include subregional differences in land cover and climate-related productivity. Reduced evidence of storm impacts by the first growing season was common the tropics, which is consistent with higher productivity and rapid vegetative recovery, but some tropical hurricane footprints showed increased departure over time. This suggests that our measures of change may capture a combination of disturbance impacts and background variation in climate or LSP that could affect our ability to recognize change.
Our results underscore the need for a strategic approach to monitoring post-hurricane impacts that is tailored to the unique ecological conditions and land-use configurations of each subregion. Our findings emphasize the value of early monitoring for the tropics where rapid sprouting can obscure structural damage. However, rapid assessments that occur during the same growing season as the storm can misrepresent minor defoliation as structural damage. Rapid refoliation can occur in areas with or without structural damage, thus confounding estimates of vegetation structural recovery. This is not to say that rapid assessments with coarse spatial resolution lack value, as they can suggest the footprint of potential impacts where more focused evaluations can be pursued. This research shows that by the first year after the storm, structural impacts are difficult to resolve in the tropics, which means that subsequent monitoring of structural recovery will be even more difficult. This affects our ability to accurately assess the dynamics and resilience of these disturbance-prone ecosystems using this technology.
Our research shows that optical techniques are more effective for assessing immediate and long-term impacts to temperate regions such as SUS. This is not just because more research has been conducted there, as analyses benefit from the subregion’s relatively predictable phenological behavior and the sensitivity of industrial pines to storm damage \[[18](https://www.mdpi.com/1999-4907/17/4/419#B18-forests-17-00419)\]. The strong seasonality of temperate latitude forests means that rapid sprouting and refoliation are less likely to be concerns here. Outside this subregion, impact and monitoring assessments are more likely to struggle with distinguishing the recovery shown by spectral indices from structural recovery. The latter, more elusive phenomenon is needed to characterize long-term impacts, ecosystem resilience, and the global implications of these frequent, large-scale storms \[[53](https://www.mdpi.com/1999-4907/17/4/419#B53-forests-17-00419)\].
Hurricane impact monitoring will benefit from understanding how and why vegetation types vary from year to year normally. Non-woody vegetation may exhibit a surge in late-season growth from the influx of tropical storm rainfall in ways that interfere with rapid forest damage assessment, but it is unclear if this short-term mechanism affects the rate of woody plant refoliation. The cross-seasonal and year-to-year productivity of non-woody types can also vary in response to drought. Thus, weather variability can affect the integrity of either the normal baseline condition or the post-disturbance indicator of damage. Where vegetation is fragmented or mixed, monitoring at high spatial resolution with forest masks can reduce this problem. Conversely, hurricane footprints with relatively homogeneous forest cover have less issues with mixed responses from fractional cover and are more amenable to assessment from coarse resolution imagery.
Better use of LSP can improve our understanding of hurricane impacts across the ecologically and climatically diverse NAB. LSP encapsulates vegetation type and productivity, the seasonal and interannual dynamics that need consideration, and the actual storm impacts. LSP is more difficult to leverage in the humid tropics where high-resolution analyses, synthetic aperture radar (SAR) and (LiDAR) technologies that resolve structural change and recovery more precisely appear to be warranted.
## Author Contributions
Conceptualization and methodology, S.P.N. and C.T.-P.; analysis and investigation, C.T.-P. and S.P.N.; data curation, C.T.-P. and S.P.N.; writing—original draft preparation, C.T.-P. and S.P.N.; writing—review and editing C.T.-P. and S.P.N.; visualization, C.T.-P. and S.P.N.; project administration: S.P.N. All authors have read and agreed to the published version of the manuscript.
## Funding
This research received no external funding.
## Data Availability Statement
The data supporting the conclusions of this article will be made available by the authors upon request.
## Acknowledgments
This research was supported in part by an appointment to the United States Forest Service (USFS) Research Participation Program administered by the Oak Ridge Institute for Science and Education (ORISE) through an interagency agreement between the U.S. Department of Energy (DOE) and the U.S. Department of Agriculture (USDA). The findings and conclusions in this publication are those of the authors and should not be construed to represent any official USDA or U.S. Government determination or policy.
## Conflicts of Interest
The authors declare that they have no commercial or associative interests that represent conflicts of interest in connection with the article submitted.
## Abbreviations
The following abbreviations are used in this manuscript:
| | |
|---|---|
| LSP | Land Surface Phenology |
| EVI | Enhanced Vegetation Index |
| AUC | Area Under the Curve |
| NAB | North Atlantic Basin |
| C | Caribbean |
| CA | Central America |
| FP | Florida Peninsula |
| SG | South Gulf |
| SUS | Southeast United States |
| WG | West Gulf |
## Appendix A
![]()
**Table A1.** Percent of land cover types within the 44 hurricanes focal areas derived from MODIS.
**Table A1.** Percent of land cover types within the 44 hurricanes focal areas derived from MODIS.
| | | | | | | | |
|---|---|---|---|---|---|---|---|
| Florida Peninsula (FP) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| FRANCES | 1\.1 | 0\.4 | 0\.4 | 6\.7 | 1\.2 | 8 | 82\.2 |
| IRMA | 0\.4 | 0\.7 | 1\.2 | 7\.6 | 12\.5 | 23\.3 | 54\.3 |
| JEANNE | 0\.8 | 0\.1 | 0\.4 | 7\.6 | 1\.7 | 11\.8 | 77\.6 |
| WILMA | 1\.9 | 0\.5 | 3\.2 | 3\.9 | 18\.7 | 34\.1 | 37\.8 |
| CHARLEY \* | 0\.2 | 0\.1 | 0\.7 | 12\.4 | 2\.9 | 9\.1 | 74\.6 |
| IAN | 0\.1 | 0\.1 | 0\.8 | 9\.9 | 3\.1 | 9 | 77\.1 |
| Caribbean (C) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| PALOMA | 31\.8 | 12 | 6\.4 | 1\.3 | 17\.9 | 1\.6 | 29 |
| RICHARD | 0\.5 | 0 | 72\.9 | 8\.8 | 12\.2 | 2\.8 | 2\.7 |
| GRACE | 8 | 0 | 5\.7 | 2\.6 | 50\.9 | 2 | 30\.9 |
| IAN | 22\.4 | 0\.1 | 17\.3 | 5\.7 | 21\.9 | 5\.4 | 27\.2 |
| SANDY | 37\.7 | 0\.3 | 18\.1 | 1\.4 | 24\.2 | 3\.6 | 14\.6 |
| DENNIS | 18\.4 | 1\.6 | 19\.6 | 1\.6 | 11\.9 | 16\.4 | 30\.5 |
| MARIA \* | 6\.9 | 0 | 25 | 7\.2 | 41\.8 | 14\.5 | 4\.6 |
| MICHELLE | 31\.6 | 7\.7 | 13\.1 | 0\.6 | 9\.5 | 6 | 31\.5 |
| WILMA | 0 | 0\.1 | 67\.3 | 3\.4 | 20\.3 | 8\.6 | 0\.3 |
| GUSTAV | 27\.9 | 0\.2 | 15\.6 | 3\.3 | 16\.5 | 5\.3 | 31\.3 |
| South Gulf (SG) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| DELTA | 0\.3 | 2 | 71\.9 | 2\.6 | 20\.4 | 1\.1 | 1\.7 |
| ERNESTO | 0\.1 | 0\.1 | 80\.5 | 2\.9 | 11\.7 | 3\.8 | 0\.9 |
| ISIDORE | 0\.3 | 32\.1 | 0 | 4\.6 | 53\.6 | 4 | 5\.5 |
| KARL | 43\.1 | 1\.8 | 14\.1 | 0\.9 | 14\.8 | 1\.8 | 23\.6 |
| EMILY \* | 0\.1 | 10\.9 | 78\.6 | 0\.6 | 9\.2 | 0\.5 | 0 |
| DEAN | 0 | 0\.2 | 81\.3 | 2\.4 | 11\.7 | 3\.7 | 0\.7 |
| Central America (CA) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| BETA | 0 | 0 | 29 | 1\.9 | 33\.5 | 3\.7 | 31\.8 |
| IRIS | 0 | 0 | 56\.4 | 6\.9 | 30\.4 | 0\.2 | 6 |
| OTTO | 0 | 0 | 52\.7 | 1\.4 | 25\.7 | 0\.9 | 19\.3 |
| ETA | 0 | 0 | 43\.6 | 9\.8 | 33\.5 | 8\.5 | 4\.6 |
| IOTA \* | 0 | 0 | 45\.2 | 4\.2 | 32\.7 | 6\.2 | 11\.7 |
| FELIX | 0 | 0 | 40\.7 | 18\.9 | 13 | 8\.1 | 19\.2 |
| Southeast U.S. (SUS) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| DELTA | 39\.4 | 3\.4 | 2\.4 | 4\.7 | 8\.7 | 15 | 26\.5 |
| GUSTAV | 2\.1 | 6 | 0\.1 | 1\.9 | 6\.2 | 75\.3 | 8\.4 |
| IKE | 0\.9 | 2\.6 | 8\.7 | 6\.4 | 39\.4 | 16\.4 | 25\.7 |
| ISABEL | 21\.3 | 6\.2 | 3\.1 | 1\.6 | 55\.9 | 9\.9 | 2\.1 |
| SALLY | 10\.8 | 3\.1 | 14\.4 | 7\.6 | 51\.1 | 3 | 10 |
| DENNIS | 5\.7 | 11\.3 | 10\.4 | 3\.8 | 62\.1 | 2\.6 | 4\.1 |
| IVAN | 4\.5 | 19\.4 | 12\.8 | 3 | 47\.4 | 7\.6 | 5\.3 |
| KATRINA | 0 | 11 | 6\.8 | 2\.2 | 66\.7 | 5\.5 | 7\.8 |
| RITA | 0\.2 | 5\.8 | 11\.5 | 2\.3 | 40\.6 | 17\.9 | 21\.7 |
| ZETA | 1\.3 | 2\.6 | 0 | 2\.7 | 4\.2 | 62\.6 | 26\.5 |
| IDA | 8\.6 | 15\.7 | 0 | 3 | 12 | 39 | 21\.7 |
| LAURA \* | 4\.1 | 2 | 13\.4 | 4\.4 | 32\.4 | 19\.1 | 24\.6 |
| MICHAEL | 4\.9 | 5\.5 | 22\.5 | 3\.6 | 52\.4 | 6 | 5\.1 |
| West Gulf (WG) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| ALEX | 0\.4 | 4\.7 | 2\.5 | 44\.5 | 6 | 5\.9 | 35\.9 |
| EMILY | 5\.1 | 0\.1 | 0 | 66\.3 | 2\.4 | 15\.7 | 10\.3 |
| HARVEY \* | 4\.4 | 0\.1 | 0 | 25\.3 | 0 | 4\.6 | 65\.7 |
\* The hurricanes examined as case studies.
## References
1. Cole, J.; Nowak, D.J.; Greenfield, E.J. Potential Hurricane Wind Risk to US Rural and Urban Forests. J. For. **2021**, 119, 393–406. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Potential+Hurricane+Wind+Risk+to+US+Rural+and+Urban+Forests&author=Cole,+J.&author=Nowak,+D.J.&author=Greenfield,+E.J.&publication_year=2021&journal=J.+For.&volume=119&pages=393%E2%80%93406&doi=10.1093/jofore/fvab018)\] \[[CrossRef](https://doi.org/10.1093/jofore/fvab018)\]
2. Mallin, M.A.; Corbett, C.A. How Hurricane Attributes Determine the Extent of Environmental Effects: Multiple Hurricanes and Different Coastal Systems. Estuaries Coasts **2006**, 29, 1046–1061. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=How+Hurricane+Attributes+Determine+the+Extent+of+Environmental+Effects:+Multiple+Hurricanes+and+Different+Coastal+Systems&author=Mallin,+M.A.&author=Corbett,+C.A.&publication_year=2006&journal=Estuaries+Coasts&volume=29&pages=1046%E2%80%931061&doi=10.1007/BF02798667)\] \[[CrossRef](https://doi.org/10.1007/BF02798667)\]
3. Reed, K.A.; Wehner, M.F.; Zarzycki, C.M. Attribution of 2020 Hurricane Season Extreme Rainfall to Human-Induced Climate Change. Nat. Commun. **2022**, 13, 1905. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Attribution+of+2020+Hurricane+Season+Extreme+Rainfall+to+Human-Induced+Climate+Change&author=Reed,+K.A.&author=Wehner,+M.F.&author=Zarzycki,+C.M.&publication_year=2022&journal=Nat.+Commun.&volume=13&pages=1905&doi=10.1038/s41467-022-29379-1&pmid=35414063)\] \[[CrossRef](https://doi.org/10.1038/s41467-022-29379-1)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/35414063)\]
4. Sharma, A.; Ojha, S.K.; Dimov, L.D.; Vogel, J.G.; Nowak, J. Long-Term Effects of Catastrophic Wind on Southern US Coastal Forests: Lessons from a Major Hurricane. PLoS ONE **2021**, 16, e0243362. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Long-Term+Effects+of+Catastrophic+Wind+on+Southern+US+Coastal+Forests:+Lessons+from+a+Major+Hurricane&author=Sharma,+A.&author=Ojha,+S.K.&author=Dimov,+L.D.&author=Vogel,+J.G.&author=Nowak,+J.&publication_year=2021&journal=PLoS+ONE&volume=16&pages=e0243362&doi=10.1371/journal.pone.0243362)\] \[[CrossRef](https://doi.org/10.1371/journal.pone.0243362)\]
5. Marsooli, R.; Lin, N.; Emanuel, K.; Feng, K. Climate Change Exacerbates Hurricane Flood Hazards along US Atlantic and Gulf Coasts in Spatially Varying Patterns. Nat. Commun. **2019**, 10, 3785. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Climate+Change+Exacerbates+Hurricane+Flood+Hazards+along+US+Atlantic+and+Gulf+Coasts+in+Spatially+Varying+Patterns&author=Marsooli,+R.&author=Lin,+N.&author=Emanuel,+K.&author=Feng,+K.&publication_year=2019&journal=Nat.+Commun.&volume=10&pages=3785&doi=10.1038/s41467-019-11755-z&pmid=31439853)\] \[[CrossRef](https://doi.org/10.1038/s41467-019-11755-z)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/31439853)\]
6. Vecchi, G.A.; Landsea, C.; Zhang, W.; Villarini, G.; Knutson, T. Changes in Atlantic Major Hurricane Frequency since the Late-19th Century. Nat. Commun. **2021**, 12, 4054. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Changes+in+Atlantic+Major+Hurricane+Frequency+since+the+Late-19th+Century&author=Vecchi,+G.A.&author=Landsea,+C.&author=Zhang,+W.&author=Villarini,+G.&author=Knutson,+T.&publication_year=2021&journal=Nat.+Commun.&volume=12&pages=4054&doi=10.1038/s41467-021-24268-5&pmid=34257285)\] \[[CrossRef](https://doi.org/10.1038/s41467-021-24268-5)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/34257285)\]
7. Lin, T.C.; Hogan, J.A.; Chang, C.T. Tropical Cyclone Ecology: A Scale-Link Perspective. Trends Ecol. Evol. **2020**, 35, 594–604. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Tropical+Cyclone+Ecology:+A+Scale-Link+Perspective&author=Lin,+T.C.&author=Hogan,+J.A.&author=Chang,+C.T.&publication_year=2020&journal=Trends+Ecol.+Evol.&volume=35&pages=594%E2%80%93604&doi=10.1016/j.tree.2020.02.012)\] \[[CrossRef](https://doi.org/10.1016/j.tree.2020.02.012)\]
8. Gao, W.; Larjavaara, M. Wind Disturbance in Forests: A Bibliometric Analysis and Systematic Review. For. Ecol. Manag. **2024**, 564, 122001. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Wind+Disturbance+in+Forests:+A+Bibliometric+Analysis+and+Systematic+Review&author=Gao,+W.&author=Larjavaara,+M.&publication_year=2024&journal=For.+Ecol.+Manag.&volume=564&pages=122001&doi=10.1016/j.foreco.2024.122001)\] \[[CrossRef](https://doi.org/10.1016/j.foreco.2024.122001)\]
9. Hernandez, J.O.; Maldia, L.S.J.; Park, B.B.; Hernandez, J.O.; Maldia, L.S.J.; Park, B.B. Research Trends and Methodological Approaches of the Impacts of Windstorms on Forests in Tropical, Subtropical, and Temperate Zones: Where Are We Now and How Should Research Move Forward? Plants **2020**, 9, 1709. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Research+Trends+and+Methodological+Approaches+of+the+Impacts+of+Windstorms+on+Forests+in+Tropical,+Subtropical,+and+Temperate+Zones:+Where+Are+We+Now+and+How+Should+Research+Move+Forward?&author=Hernandez,+J.O.&author=Maldia,+L.S.J.&author=Park,+B.B.&author=Hernandez,+J.O.&author=Maldia,+L.S.J.&author=Park,+B.B.&publication_year=2020&journal=Plants&volume=9&pages=1709&doi=10.3390/plants9121709&pmid=33291785)\] \[[CrossRef](https://doi.org/10.3390/plants9121709)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/33291785)\]
10. Henebry, G.M.; de Beurs, K.M. Remote Sensing of Land Surface Phenology: A Prospectus. In Phenology: An Integrative Environmental Science; Schwartz, M.D., Ed.; Springer: Dordrecht, The Netherlands, 2013; pp. 385–411. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Remote+Sensing+of+Land+Surface+Phenology:+A+Prospectus&author=Henebry,+G.M.&author=de+Beurs,+K.M.&publication_year=2013&pages=385%E2%80%93411)\]
11. Pomara, L.Y.; Lee, D.C.; Brooks, B.-G.; Hargrove, W.W. Using Land Surface Phenology and Information Theory to Assess and Map Complex Landscape Dynamics. Landsc. Ecol. **2024**, 39, 203. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Using+Land+Surface+Phenology+and+Information+Theory+to+Assess+and+Map+Complex+Landscape+Dynamics&author=Pomara,+L.Y.&author=Lee,+D.C.&author=Brooks,+B.-G.&author=Hargrove,+W.W.&publication_year=2024&journal=Landsc.+Ecol.&volume=39&pages=203&doi=10.1007/s10980-024-02005-9)\] \[[CrossRef](https://doi.org/10.1007/s10980-024-02005-9)\]
12. White, M.A.; de Beurs, K.M.; Didan, K.; Inouye, D.W.; Richardson, A.D.; Jensen, O.P.; O’Keefe, J.; Zhang, G.; Nemani, R.R.; van Leeuwen, W.J.D.; et al. Intercomparison, Interpretation, and Assessment of Spring Phenology in North America Estimated from Remote Sensing for 1982-2006. Glob. Change Biol. **2009**, 15, 2335–2359. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Intercomparison,+Interpretation,+and+Assessment+of+Spring+Phenology+in+North+America+Estimated+from+Remote+Sensing+for+1982-2006&author=White,+M.A.&author=de+Beurs,+K.M.&author=Didan,+K.&author=Inouye,+D.W.&author=Richardson,+A.D.&author=Jensen,+O.P.&author=O%E2%80%99Keefe,+J.&author=Zhang,+G.&author=Nemani,+R.R.&author=van+Leeuwen,+W.J.D.&author=et+al.&publication_year=2009&journal=Glob.+Change+Biol.&volume=15&pages=2335%E2%80%932359&doi=10.1111/j.1365-2486.2009.01910.x)\] \[[CrossRef](https://doi.org/10.1111/j.1365-2486.2009.01910.x)\]
13. Caparros-Santiago, J.A.; Rodriguez-Galiano, V.; Dash, J. Land Surface Phenology as Indicator of Global Terrestrial Ecosystem Dynamics: A Systematic Review. ISPRS J. Photogramm. Remote Sens. **2021**, 171, 330–347. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Land+Surface+Phenology+as+Indicator+of+Global+Terrestrial+Ecosystem+Dynamics:+A+Systematic+Review&author=Caparros-Santiago,+J.A.&author=Rodriguez-Galiano,+V.&author=Dash,+J.&publication_year=2021&journal=ISPRS+J.+Photogramm.+Remote+Sens.&volume=171&pages=330%E2%80%93347&doi=10.1016/j.isprsjprs.2020.11.019)\] \[[CrossRef](https://doi.org/10.1016/j.isprsjprs.2020.11.019)\]
14. Kooistra, L.; Berger, K.; Brede, B.; Graf, L.V.; Aasen, H.; Roujean, J.L.; Machwitz, M.; Schlerf, M.; Atzberger, C.; Prikaziuk, E.; et al. Reviews and Syntheses: Remotely Sensed Optical Time Series for Monitoring Vegetation Productivity. Biogeosciences **2024**, 21, 473–511. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Reviews+and+Syntheses:+Remotely+Sensed+Optical+Time+Series+for+Monitoring+Vegetation+Productivity&author=Kooistra,+L.&author=Berger,+K.&author=Brede,+B.&author=Graf,+L.V.&author=Aasen,+H.&author=Roujean,+J.L.&author=Machwitz,+M.&author=Schlerf,+M.&author=Atzberger,+C.&author=Prikaziuk,+E.&author=et+al.&publication_year=2024&journal=Biogeosciences&volume=21&pages=473%E2%80%93511&doi=10.5194/bg-21-473-2024)\] \[[CrossRef](https://doi.org/10.5194/bg-21-473-2024)\]
15. Cui, K.L.; Yang, J.L.; Dong, J.W.; Zhao, G.S.; Cui, Y.P. Comparing Different Spatial Resolutions and Indices for Retrieving Land Surface Phenology for Deciduous Broadleaf Forests. Remote Sens. **2023**, 15, 2266. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Comparing+Different+Spatial+Resolutions+and+Indices+for+Retrieving+Land+Surface+Phenology+for+Deciduous+Broadleaf+Forests&author=Cui,+K.L.&author=Yang,+J.L.&author=Dong,+J.W.&author=Zhao,+G.S.&author=Cui,+Y.P.&publication_year=2023&journal=Remote+Sens.&volume=15&pages=2266&doi=10.3390/rs15092266)\] \[[CrossRef](https://doi.org/10.3390/rs15092266)\]
16. Gnilke, A.; Sanders, T.G.M. Distinguishing Abrupt and Gradual Forest Disturbances with MODIS-Based Phenological Anomaly Series. Front. Plant Sci. **2022**, 13, 863116. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Distinguishing+Abrupt+and+Gradual+Forest+Disturbances+with+MODIS-Based+Phenological+Anomaly+Series&author=Gnilke,+A.&author=Sanders,+T.G.M.&publication_year=2022&journal=Front.+Plant+Sci.&volume=13&pages=863116&doi=10.3389/fpls.2022.863116)\] \[[CrossRef](https://doi.org/10.3389/fpls.2022.863116)\]
17. Guzmán Q., J.A.; Pinto-Ledezma, J.N.; Frantz, D.; Townsend, P.A.; Juzwik, J.; Cavender-Bares, J. Mapping Oak Wilt Disease from Space Using Land Surface Phenology. Remote Sens. Environ. **2023**, 298, 113794. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Mapping+Oak+Wilt+Disease+from+Space+Using+Land+Surface+Phenology&author=Guzm%C3%A1n+Q.,+J.A.&author=Pinto-Ledezma,+J.N.&author=Frantz,+D.&author=Townsend,+P.A.&author=Juzwik,+J.&author=Cavender-Bares,+J.&publication_year=2023&journal=Remote+Sens.+Environ.&volume=298&pages=113794&doi=10.1016/j.rse.2023.113794)\] \[[CrossRef](https://doi.org/10.1016/j.rse.2023.113794)\]
18. Topete-Pozas, C.; Norman, S.P.; Christie, W.M. Multi-Year Hurricane Impacts Across an Urban-to-Industrial Forest Use Gradient. Remote Sens. **2024**, 16, 3890. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Multi-Year+Hurricane+Impacts+Across+an+Urban-to-Industrial+Forest+Use+Gradient&author=Topete-Pozas,+C.&author=Norman,+S.P.&author=Christie,+W.M.&publication_year=2024&journal=Remote+Sens.&volume=16&pages=3890&doi=10.3390/rs16203890)\] \[[CrossRef](https://doi.org/10.3390/rs16203890)\]
19. Wang, F.G.; Xu, Y.J. Hurricane Katrina-Induced Forest Damage in Relation to Ecological Factors at Landscape Scale. Environ. Monit. Assess. **2009**, 156, 491–507. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Hurricane+Katrina-Induced+Forest+Damage+in+Relation+to+Ecological+Factors+at+Landscape+Scale&author=Wang,+F.G.&author=Xu,+Y.J.&publication_year=2009&journal=Environ.+Monit.+Assess.&volume=156&pages=491%E2%80%93507&doi=10.1007/s10661-008-0500-6)\] \[[CrossRef](https://doi.org/10.1007/s10661-008-0500-6)\]
20. Wang, W.T.; Qu, J.J.; Hao, X.J.; Liu, Y.Q.; Stanturf, J.A. Post-Hurricane Forest Damage Assessment Using Satellite Remote Sensing. Agric. For. Meteorol. **2010**, 150, 122–132. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Post-Hurricane+Forest+Damage+Assessment+Using+Satellite+Remote+Sensing&author=Wang,+W.T.&author=Qu,+J.J.&author=Hao,+X.J.&author=Liu,+Y.Q.&author=Stanturf,+J.A.&publication_year=2010&journal=Agric.+For.+Meteorol.&volume=150&pages=122%E2%80%93132&doi=10.1016/j.agrformet.2009.09.009)\] \[[CrossRef](https://doi.org/10.1016/j.agrformet.2009.09.009)\]
21. Feng, Y.L.; Negron-Juarez, R.I.; Chambers, J.Q. Remote Sensing and Statistical Analysis of the Effects of Hurricane Maria on the Forests of Puerto Rico. Remote Sens. Environ. **2020**, 247, 111940. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Remote+Sensing+and+Statistical+Analysis+of+the+Effects+of+Hurricane+Maria+on+the+Forests+of+Puerto+Rico&author=Feng,+Y.L.&author=Negron-Juarez,+R.I.&author=Chambers,+J.Q.&publication_year=2020&journal=Remote+Sens.+Environ.&volume=247&pages=111940&doi=10.1016/j.rse.2020.111940)\] \[[CrossRef](https://doi.org/10.1016/j.rse.2020.111940)\]
22. Dupuis, C.; Lejeune, P.; Michez, A.; Fayolle, A. How Can Remote Sensing Help Monitor Tropical Moist Forest Degradation?—A Systematic Review. Remote Sens. **2020**, 12, 1087. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=How+Can+Remote+Sensing+Help+Monitor+Tropical+Moist+Forest+Degradation?%E2%80%94A+Systematic+Review&author=Dupuis,+C.&author=Lejeune,+P.&author=Michez,+A.&author=Fayolle,+A.&publication_year=2020&journal=Remote+Sens.&volume=12&pages=1087&doi=10.3390/rs12071087)\] \[[CrossRef](https://doi.org/10.3390/rs12071087)\]
23. Helman, D. Land Surface Phenology: What Do We Really “see” from Space? Sci. Total Environ. **2018**, 618, 665–673. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Land+Surface+Phenology:+What+Do+We+Really+%E2%80%9Csee%E2%80%9D+from+Space?&author=Helman,+D.&publication_year=2018&journal=Sci.+Total+Environ.&volume=618&pages=665%E2%80%93673&doi=10.1016/j.scitotenv.2017.07.237)\] \[[CrossRef](https://doi.org/10.1016/j.scitotenv.2017.07.237)\]
24. Huete, A.; Didan, K.; Miura, T.; Rodriguez, E.P.; Gao, X.; Ferreira, L.G. Overview of the Radiometric and Biophysical Performance of the MODIS Vegetation Indices. Remote Sens. Environ. **2002**, 83, 195–213. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Overview+of+the+Radiometric+and+Biophysical+Performance+of+the+MODIS+Vegetation+Indices&author=Huete,+A.&author=Didan,+K.&author=Miura,+T.&author=Rodriguez,+E.P.&author=Gao,+X.&author=Ferreira,+L.G.&publication_year=2002&journal=Remote+Sens.+Environ.&volume=83&pages=195%E2%80%93213&doi=10.1016/S0034-4257\(02\)00096-2)\] \[[CrossRef](https://doi.org/10.1016/S0034-4257\(02\)00096-2)\]
25. Rogan, J.; Schneider, L.; Christman, Z.; Millones, M.; Lawrence, D.; Schmook, B. Hurricane Disturbance Mapping Using MODIS EVI Data in the Southeastern Yucatan, Mexico. Remote Sens. Lett. **2011**, 2, 259–267. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Hurricane+Disturbance+Mapping+Using+MODIS+EVI+Data+in+the+Southeastern+Yucatan,+Mexico&author=Rogan,+J.&author=Schneider,+L.&author=Christman,+Z.&author=Millones,+M.&author=Lawrence,+D.&author=Schmook,+B.&publication_year=2011&journal=Remote+Sens.+Lett.&volume=2&pages=259%E2%80%93267&doi=10.1080/01431161.2010.520344)\] \[[CrossRef](https://doi.org/10.1080/01431161.2010.520344)\]
26. Rossi, E.; Rogan, J.; Schneider, L. Mapping Forest Damage in Northern Nicaragua after Hurricane Felix (2007) Using MODIS Enhanced Vegetation Index Data. Gisci. Remote Sens. **2013**, 50, 385–399. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Mapping+Forest+Damage+in+Northern+Nicaragua+after+Hurricane+Felix+\(2007\)+Using+MODIS+Enhanced+Vegetation+Index+Data&author=Rossi,+E.&author=Rogan,+J.&author=Schneider,+L.&publication_year=2013&journal=Gisci.+Remote+Sens.&volume=50&pages=385%E2%80%93399&doi=10.1080/15481603.2013.820066)\] \[[CrossRef](https://doi.org/10.1080/15481603.2013.820066)\]
27. Wang, F.G.; D’Sa, E.J. Potential of MODIS EVI in Identifying Hurricane Disturbance to Coastal Vegetation in the Northern Gulf of Mexico. Remote Sens. **2010**, 2, 1–18. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Potential+of+MODIS+EVI+in+Identifying+Hurricane+Disturbance+to+Coastal+Vegetation+in+the+Northern+Gulf+of+Mexico&author=Wang,+F.G.&author=D%E2%80%99Sa,+E.J.&publication_year=2010&journal=Remote+Sens.&volume=2&pages=1%E2%80%9318&doi=10.3390/rs2010001)\] \[[CrossRef](https://doi.org/10.3390/rs2010001)\]
28. Gang, C.C.; Pan, S.F.; Tian, H.Q.; Wang, Z.N.; Xu, R.T.; Bian, Z.H.; Pan, N.Q.; Yao, Y.Z.; Shi, H. Satellite Observations of Forest Resilience to Hurricanes along the Northern Gulf of Mexico. For. Ecol. Manag. **2020**, 472, 118243. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Satellite+Observations+of+Forest+Resilience+to+Hurricanes+along+the+Northern+Gulf+of+Mexico&author=Gang,+C.C.&author=Pan,+S.F.&author=Tian,+H.Q.&author=Wang,+Z.N.&author=Xu,+R.T.&author=Bian,+Z.H.&author=Pan,+N.Q.&author=Yao,+Y.Z.&author=Shi,+H.&publication_year=2020&journal=For.+Ecol.+Manag.&volume=472&pages=118243&doi=10.1016/j.foreco.2020.118243)\] \[[CrossRef](https://doi.org/10.1016/j.foreco.2020.118243)\]
29. de Beurs, K.M.; McThompson, N.S.; Owsley, B.C.; Henebry, G.M. Hurricane Damage Detection on Four Major Caribbean Islands. Remote Sens. Environ. **2019**, 229, 1–13. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Hurricane+Damage+Detection+on+Four+Major+Caribbean+Islands&author=de+Beurs,+K.M.&author=McThompson,+N.S.&author=Owsley,+B.C.&author=Henebry,+G.M.&publication_year=2019&journal=Remote+Sens.+Environ.&volume=229&pages=1%E2%80%9313&doi=10.1016/j.rse.2019.04.028)\] \[[CrossRef](https://doi.org/10.1016/j.rse.2019.04.028)\]
30. Turner, H.C.; Galford, G.L.; Lopez, N.H.; Méndez, A.F.; Borroto-Escuela, D.Y.; Ramos, I.H.; González-Díaz, P. Extent, Severity, and Temporal Patterns of Damage to Cuba’s Ecosystems Following Hurricane Irma: MODIS and Sentinel-2 Hurricane Disturbance Vegetation Anomaly (HDVA). Remote Sens. **2023**, 15, 2495. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Extent,+Severity,+and+Temporal+Patterns+of+Damage+to+Cuba%E2%80%99s+Ecosystems+Following+Hurricane+Irma:+MODIS+and+Sentinel-2+Hurricane+Disturbance+Vegetation+Anomaly+\(HDVA\)&author=Turner,+H.C.&author=Galford,+G.L.&author=Lopez,+N.H.&author=M%C3%A9ndez,+A.F.&author=Borroto-Escuela,+D.Y.&author=Ramos,+I.H.&author=Gonz%C3%A1lez-D%C3%ADaz,+P.&publication_year=2023&journal=Remote+Sens.&volume=15&pages=2495&doi=10.3390/rs15102495)\] \[[CrossRef](https://doi.org/10.3390/rs15102495)\]
31. Dronova, I.; Taddeo, S. Remote Sensing of Phenology: Towards the Comprehensive Indicators of Plant Community Dynamics from Species to Regional Scales. J. Ecol. **2022**, 110, 1460–1484. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Remote+Sensing+of+Phenology:+Towards+the+Comprehensive+Indicators+of+Plant+Community+Dynamics+from+Species+to+Regional+Scales&author=Dronova,+I.&author=Taddeo,+S.&publication_year=2022&journal=J.+Ecol.&volume=110&pages=1460%E2%80%931484&doi=10.1111/1365-2745.13897)\] \[[CrossRef](https://doi.org/10.1111/1365-2745.13897)\]
32. Hogan, J.A.; Feagin, R.A.; Starr, G.; Ross, M.; Lin, T.C.; O’connell, C.; Huff, T.P.; Stauffer, B.A.; Robinson, K.L.; Lara, M.C.; et al. A Research Framework to Integrate Cross-Ecosystem Responses to Tropical Cyclones. Bioscience **2020**, 70, 477–489. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=A+Research+Framework+to+Integrate+Cross-Ecosystem+Responses+to+Tropical+Cyclones&author=Hogan,+J.A.&author=Feagin,+R.A.&author=Starr,+G.&author=Ross,+M.&author=Lin,+T.C.&author=O%E2%80%99connell,+C.&author=Huff,+T.P.&author=Stauffer,+B.A.&author=Robinson,+K.L.&author=Lara,+M.C.&author=et+al.&publication_year=2020&journal=Bioscience&volume=70&pages=477%E2%80%93489&doi=10.1093/biosci/biaa034)\] \[[CrossRef](https://doi.org/10.1093/biosci/biaa034)\]
33. Kenneth, R.; Howard, J.; James, P.; Michael, C.; Carl, J. International Best Track Archive for Climate Stewardship (IBTrACS) Project; Version 4; NOAA National Centers for Environmental Information (NCEI): Asheville, NC, USA, 2019. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=International+Best+Track+Archive+for+Climate+Stewardship+\(IBTrACS\)+Project&author=Kenneth,+R.&author=Howard,+J.&author=James,+P.&author=Michael,+C.&author=Carl,+J.&publication_year=2019)\]
34. Zhou, Y.; Matyas, C.J. Regionalization of Precipitation Associated with Tropical Cyclones Using Spatial Metrics and Satellite Precipitation. GISci. Remote Sens. **2021**, 58, 542–561. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Regionalization+of+Precipitation+Associated+with+Tropical+Cyclones+Using+Spatial+Metrics+and+Satellite+Precipitation&author=Zhou,+Y.&author=Matyas,+C.J.&publication_year=2021&journal=GISci.+Remote+Sens.&volume=58&pages=542%E2%80%93561&doi=10.1080/15481603.2021.1908675)\] \[[CrossRef](https://doi.org/10.1080/15481603.2021.1908675)\]
35. Kim, D.; Matyas, C.J. Classification of Tropical Cyclone Rain Patterns Using Convolutional Autoencoder. Sci. Rep. **2024**, 14, 791. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Classification+of+Tropical+Cyclone+Rain+Patterns+Using+Convolutional+Autoencoder&author=Kim,+D.&author=Matyas,+C.J.&publication_year=2024&journal=Sci.+Rep.&volume=14&pages=791&doi=10.1038/s41598-023-50994-5)\] \[[CrossRef](https://doi.org/10.1038/s41598-023-50994-5)\]
36. Abatzoglou, J.T.; Dobrowski, S.Z.; Parks, S.A.; Hegewisch, K.C. TerraClimate, a High-Resolution Global Dataset of Monthly Climate and Climatic Water Balance from 1958–2015. Sci. Data **2018**, 5, 170191. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=TerraClimate,+a+High-Resolution+Global+Dataset+of+Monthly+Climate+and+Climatic+Water+Balance+from+1958%E2%80%932015&author=Abatzoglou,+J.T.&author=Dobrowski,+S.Z.&author=Parks,+S.A.&author=Hegewisch,+K.C.&publication_year=2018&journal=Sci.+Data&volume=5&pages=170191&doi=10.1038/sdata.2017.191&pmid=29313841)\] \[[CrossRef](https://doi.org/10.1038/sdata.2017.191)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/29313841)\]
37. Hargrove, W.W.; Hoffman, F.M. Using Multivariate Clustering to Characterize Ecoregion Borders. Comput. Sci. Eng. **1999**, 1, 18–25. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Using+Multivariate+Clustering+to+Characterize+Ecoregion+Borders&author=Hargrove,+W.W.&author=Hoffman,+F.M.&publication_year=1999&journal=Comput.+Sci.+Eng.&volume=1&pages=18%E2%80%9325&doi=10.1109/5992.774837)\] \[[CrossRef](https://doi.org/10.1109/5992.774837)\]
38. R Core Team. R: A Language and Environment for Statistical Computing; R Core Team: Vienna, Austria, 2024. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=R:+A+Language+and+Environment+for+Statistical+Computing&author=R+Core+Team&publication_year=2024)\]
39. Stekhoven, D.J. missForest: Nonparametric Missing Value Imputation Using Random Forest; Harvard University: Cambridge, MA, USA, 2025. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=missForest:+Nonparametric+Missing+Value+Imputation+Using+Random+Forest&author=Stekhoven,+D.J.&publication_year=2025)\]
40. Kong, D.D.; Zhang, Y.Q.; Gu, X.H.; Wang, D.G. A Robust Method for Reconstructing Global MODIS EVI Time Series on the Google Earth Engine. ISPRS J. Photogramm. Remote Sens. **2019**, 155, 13–24. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=A+Robust+Method+for+Reconstructing+Global+MODIS+EVI+Time+Series+on+the+Google+Earth+Engine&author=Kong,+D.D.&author=Zhang,+Y.Q.&author=Gu,+X.H.&author=Wang,+D.G.&publication_year=2019&journal=ISPRS+J.+Photogramm.+Remote+Sens.&volume=155&pages=13%E2%80%9324&doi=10.1016/j.isprsjprs.2019.06.014)\] \[[CrossRef](https://doi.org/10.1016/j.isprsjprs.2019.06.014)\]
41. Savitzky, A.; Golay, M.J.E. Smoothing + Differentiation of Data by Simplified Least Squares Procedures. Anal. Chem. **1964**, 36, 1627–1639. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Smoothing+++Differentiation+of+Data+by+Simplified+Least+Squares+Procedures&author=Savitzky,+A.&author=Golay,+M.J.E.&publication_year=1964&journal=Anal.+Chem.&volume=36&pages=1627%E2%80%931639&doi=10.1021/ac60214a047)\] \[[CrossRef](https://doi.org/10.1021/ac60214a047)\]
42. Tian, J.Q.; Zhu, X.L.; Chen, J.; Wang, C.; Shen, M.G.; Yang, W.; Tan, X.Y.; Xu, S.; Li, Z.L. Improving the Accuracy of Spring Phenology Detection by Optimally Smoothing Satellite Vegetation Index Time Series Based on Local Cloud Frequency. ISPRS J. Photogramm. Remote Sens. **2021**, 180, 29–44. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Improving+the+Accuracy+of+Spring+Phenology+Detection+by+Optimally+Smoothing+Satellite+Vegetation+Index+Time+Series+Based+on+Local+Cloud+Frequency&author=Tian,+J.Q.&author=Zhu,+X.L.&author=Chen,+J.&author=Wang,+C.&author=Shen,+M.G.&author=Yang,+W.&author=Tan,+X.Y.&author=Xu,+S.&author=Li,+Z.L.&publication_year=2021&journal=ISPRS+J.+Photogramm.+Remote+Sens.&volume=180&pages=29%E2%80%9344&doi=10.1016/j.isprsjprs.2021.08.003)\] \[[CrossRef](https://doi.org/10.1016/j.isprsjprs.2021.08.003)\]
43. Sánchez-Rivera, G.; Gómez-Mendoza, L. Tropical Cyclone Effects on Vegetation Resilience in the Yucatan Peninsula, México, between 2000–2012. Investig. Geográficas **2022**, 203, 203–220. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Tropical+Cyclone+Effects+on+Vegetation+Resilience+in+the+Yucatan+Peninsula,+M%C3%A9xico,+between+2000%E2%80%932012&author=S%C3%A1nchez-Rivera,+G.&author=G%C3%B3mez-Mendoza,+L.&publication_year=2022&journal=Investig.+Geogr%C3%A1ficas&volume=203&pages=203%E2%80%93220&doi=10.14198/INGEO.18499)\] \[[CrossRef](https://doi.org/10.14198/INGEO.18499)\]
44. Basnet, K. Recovery of a Tropical Rain Forest after Hurricane Damage. Vegetatio **1993**, 109, 1–4. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Recovery+of+a+Tropical+Rain+Forest+after+Hurricane+Damage&author=Basnet,+K.&publication_year=1993&journal=Vegetatio&volume=109&pages=1%E2%80%934&doi=10.1007/BF00149540)\] \[[CrossRef](https://doi.org/10.1007/BF00149540)\]
45. Taylor, B.N.; Stedman, E.; Van Bloem, S.J.; Whitmire, S.L.; DeWalt, S.J. Widespread Stem Snapping but Limited Mortality Caused by a Category 5 Hurricane on the Caribbean Island of Dominica. For. Ecol. Manag. **2023**, 532, 120833. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Widespread+Stem+Snapping+but+Limited+Mortality+Caused+by+a+Category+5+Hurricane+on+the+Caribbean+Island+of+Dominica&author=Taylor,+B.N.&author=Stedman,+E.&author=Van+Bloem,+S.J.&author=Whitmire,+S.L.&author=DeWalt,+S.J.&publication_year=2023&journal=For.+Ecol.+Manag.&volume=532&pages=120833&doi=10.1016/j.foreco.2023.120833)\] \[[CrossRef](https://doi.org/10.1016/j.foreco.2023.120833)\]
46. Johnsen, K.H.; Butnor, J.R.; Kush, J.S.; Schmidtling, R.C.; Nelson, C.D. Hurricane Katrina Winds Damaged Longleaf Pine Less than Loblolly Pine. South. J. Appl. For. **2009**, 33, 178–181. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Hurricane+Katrina+Winds+Damaged+Longleaf+Pine+Less+than+Loblolly+Pine&author=Johnsen,+K.H.&author=Butnor,+J.R.&author=Kush,+J.S.&author=Schmidtling,+R.C.&author=Nelson,+C.D.&publication_year=2009&journal=South.+J.+Appl.+For.&volume=33&pages=178%E2%80%93181&doi=10.1093/sjaf/33.4.178)\] \[[CrossRef](https://doi.org/10.1093/sjaf/33.4.178)\]
47. Leitold, V.; Morton, D.C.; Martinuzzi, S.; Paynter, I.; Uriarte, M.; Keller, M.; Ferraz, A.; Cook, B.D.; Corp, L.A.; González, G. Tracking the Rates and Mechanisms of Canopy Damage and Recovery Following Hurricane Maria Using Multitemporal Lidar Data. Ecosystems **2022**, 25, 892–910. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Tracking+the+Rates+and+Mechanisms+of+Canopy+Damage+and+Recovery+Following+Hurricane+Maria+Using+Multitemporal+Lidar+Data&author=Leitold,+V.&author=Morton,+D.C.&author=Martinuzzi,+S.&author=Paynter,+I.&author=Uriarte,+M.&author=Keller,+M.&author=Ferraz,+A.&author=Cook,+B.D.&author=Corp,+L.A.&author=Gonz%C3%A1lez,+G.&publication_year=2022&journal=Ecosystems&volume=25&pages=892%E2%80%93910&doi=10.1007/s10021-021-00688-8)\] \[[CrossRef](https://doi.org/10.1007/s10021-021-00688-8)\]
48. Amaral, C.; Poulter, B.; Lagomasino, D.; Fatoyinbo, T.; Taillie, P.; Lizcano, G.; Canty, S.; Silveira, J.A.H.; Teutli-Hernández, C.; Cifuentes-Jara, M.; et al. Drivers of Mangrove Vulnerability and Resilience to Tropical Cyclones in the North Atlantic Basin. Sci. Total Environ. **2023**, 898, 165413. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Drivers+of+Mangrove+Vulnerability+and+Resilience+to+Tropical+Cyclones+in+the+North+Atlantic+Basin&author=Amaral,+C.&author=Poulter,+B.&author=Lagomasino,+D.&author=Fatoyinbo,+T.&author=Taillie,+P.&author=Lizcano,+G.&author=Canty,+S.&author=Silveira,+J.A.H.&author=Teutli-Hern%C3%A1ndez,+C.&author=Cifuentes-Jara,+M.&author=et+al.&publication_year=2023&journal=Sci.+Total+Environ.&volume=898&pages=165413&doi=10.1016/j.scitotenv.2023.165413)\] \[[CrossRef](https://doi.org/10.1016/j.scitotenv.2023.165413)\]
49. Gong, Y.; Staudhammer, C.L.; Kenney, G.; Wiesner, S.; Zhang, Y.L.; Starr, G. Vegetation Structure Drives Forest Phenological Recovery after Hurricane. Sci. Total Environ. **2021**, 774, 145651. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Vegetation+Structure+Drives+Forest+Phenological+Recovery+after+Hurricane&author=Gong,+Y.&author=Staudhammer,+C.L.&author=Kenney,+G.&author=Wiesner,+S.&author=Zhang,+Y.L.&author=Starr,+G.&publication_year=2021&journal=Sci.+Total+Environ.&volume=774&pages=145651&doi=10.1016/j.scitotenv.2021.145651)\] \[[CrossRef](https://doi.org/10.1016/j.scitotenv.2021.145651)\]
50. Munson, S.M.; Long, A.L. Climate Drives Shifts in Grass Reproductive Phenology across the Western USA. New Phytol. **2017**, 213, 1945–1955. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Climate+Drives+Shifts+in+Grass+Reproductive+Phenology+across+the+Western+USA&author=Munson,+S.M.&author=Long,+A.L.&publication_year=2017&journal=New+Phytol.&volume=213&pages=1945%E2%80%931955&doi=10.1111/nph.14327&pmid=27870060)\] \[[CrossRef](https://doi.org/10.1111/nph.14327)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/27870060)\]
51. Norman, S.P.; Koch, F.H.; Hargrove, W.W. Review of Broad-Scale Drought Monitoring of Forests: Toward an Integrated Data Mining Approach. For. Ecol. Manag. **2016**, 380, 346–358. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Review+of+Broad-Scale+Drought+Monitoring+of+Forests:+Toward+an+Integrated+Data+Mining+Approach&author=Norman,+S.P.&author=Koch,+F.H.&author=Hargrove,+W.W.&publication_year=2016&journal=For.+Ecol.+Manag.&volume=380&pages=346%E2%80%93358&doi=10.1016/j.foreco.2016.06.027)\] \[[CrossRef](https://doi.org/10.1016/j.foreco.2016.06.027)\]
52. Liang, X.; Dai, Z.; Mei, X.; Wang, R.; Zeng, W.; Fagherazzi, S. Hurricanes Induced Irreversible Large-Scale Loss of Mangrove Forests. Geophys. Res. Lett. **2025**, 52, e2025GL115692. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Hurricanes+Induced+Irreversible+Large-Scale+Loss+of+Mangrove+Forests&author=Liang,+X.&author=Dai,+Z.&author=Mei,+X.&author=Wang,+R.&author=Zeng,+W.&author=Fagherazzi,+S.&publication_year=2025&journal=Geophys.+Res.+Lett.&volume=52&pages=e2025GL115692&doi=10.1029/2025GL115692)\] \[[CrossRef](https://doi.org/10.1029/2025GL115692)\]
53. Dahal, D.; Liu, S.; Oeding, J. The Carbon Cycle and Hurricanes in the United States between 1900 and 2011. Sci. Rep. **2014**, 4, 5197. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=The+Carbon+Cycle+and+Hurricanes+in+the+United+States+between+1900+and+2011&author=Dahal,+D.&author=Liu,+S.&author=Oeding,+J.&publication_year=2014&journal=Sci.+Rep.&volume=4&pages=5197&doi=10.1038/srep05197)\] \[[CrossRef](https://doi.org/10.1038/srep05197)\]
![Forests 17 00419 g001]()
**Figure 1.** Flowchart showing the data sources (in gray) and methodological workflow.
**Figure 1.** Flowchart showing the data sources (in gray) and methodological workflow.
![Forests 17 00419 g001]()
![Forests 17 00419 g002]()
**Figure 2.** The NAB study area showing the 44 hurricane tracks. The rectangles are the focal areas used in the analyses and the colors show the results of clustering.
**Figure 2.** The NAB study area showing the 44 hurricane tracks. The rectangles are the focal areas used in the analyses and the colors show the results of clustering.
![Forests 17 00419 g002]()
![Forests 17 00419 g003]()
**Figure 3.** The fraction of land cover types for the combined hurricane focal areas for each subregion. For detailed land cover by hurricane, see [Table A1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t0A1). Region labels are the same as in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
**Figure 3.** The fraction of land cover types for the combined hurricane focal areas for each subregion. For detailed land cover by hurricane, see [Table A1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t0A1). Region labels are the same as in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
![Forests 17 00419 g003]()
![Forests 17 00419 g004]()
**Figure 4.** Regional inter-seasonal and interannual vegetation and climate variation over the full period of record 2001–2022 for the hurricanes included in each subregion. The bold line shows the median value and the boxes show central distribution of the data: (**a**) EVI, (**b**) precipitation, (**c**) temperature, (**d**) vapor pressure deficit.
**Figure 4.** Regional inter-seasonal and interannual vegetation and climate variation over the full period of record 2001–2022 for the hurricanes included in each subregion. The bold line shows the median value and the boxes show central distribution of the data: (**a**) EVI, (**b**) precipitation, (**c**) temperature, (**d**) vapor pressure deficit.
![Forests 17 00419 g004]()
![Forests 17 00419 g005]()
**Figure 5.** The timing of hurricanes with respect to annual land surface phenology (LSP). The curved lines show the 22-year median and first standard deviations of EVI for the hurricane focal areas included in each subregion. Vertical lines show the landfall dates for the hurricanes included in each subregion. The colors are the same as used in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
**Figure 5.** The timing of hurricanes with respect to annual land surface phenology (LSP). The curved lines show the 22-year median and first standard deviations of EVI for the hurricane focal areas included in each subregion. Vertical lines show the landfall dates for the hurricanes included in each subregion. The colors are the same as used in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
![Forests 17 00419 g005]()
![Forests 17 00419 g006]()
**Figure 6.** Percent change in EVI AUC compared to the median wind speed of the hurricane’s focal areas by subregion for (**a**) the first winter after the storm, and (**b**) the first year after the storm. The colors are the same as used in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
**Figure 6.** Percent change in EVI AUC compared to the median wind speed of the hurricane’s focal areas by subregion for (**a**) the first winter after the storm, and (**b**) the first year after the storm. The colors are the same as used in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
![Forests 17 00419 g006]()
![Forests 17 00419 g007]()
**Figure 7.** Five-year forest (green) and all lands (gray) LSP median EVI profiles for the six case studies centered on the year of the hurricane for (**a**) Charley, (**b**) Laura, (**c**) Harvey, (**d**) Iota, (**e**) Emily, and (**f**) Maria. Landfall dates are shown by red dashed lines.
**Figure 7.** Five-year forest (green) and all lands (gray) LSP median EVI profiles for the six case studies centered on the year of the hurricane for (**a**) Charley, (**b**) Laura, (**c**) Harvey, (**d**) Iota, (**e**) Emily, and (**f**) Maria. Landfall dates are shown by red dashed lines.
![Forests 17 00419 g007]()
![]()
**Table 1.** The 44 hurricanes analyzed in this study by subregion. The six storms selected for in-depth analysis as case studies are denoted with an \*. The wind speed corresponds to the median value within the respective track/focal area.
**Table 1.** The 44 hurricanes analyzed in this study by subregion. The six storms selected for in-depth analysis as case studies are denoted with an \*. The wind speed corresponds to the median value within the respective track/focal area.
| | | | | | | | |
|---|---|---|---|---|---|---|---|
| Florida Peninsula (FP) | Central America (CA) | | | | | | |
| Name | Category | Wind speed (km/h) | Landfall | Name | Category | Wind speed (km/h) | Landfall |
| FRANCES | 2 | 157\.42 | 2 September 2004 | BETA | 2 | 133\.34 | 30 October 2005 |
| IRMA | 3 | 178\.71 | 6 September 2017 | IRIS | 3 | 200 | 9 October 2001 |
| JEANNE | 3 | 185\.2 | 16 September 2004 | OTTO | 3 | 180\.57 | 24 November 2016 |
| WILMA | 3 | 185\.2 | 24 October 2005 | ETA | 4 | 175\.94 | 3 November 2020 |
| CHARLEY \* | 4 | 231\.5 | 13 August 2004 | IOTA \* | 4 | 203\.72 | 17 November 2020 |
| IAN | 4 | 207\.42 | 28 September 2022 | FELIX | 5 | 233\.35 | 4 September 2007 |
| Caribbean (C) | Southeast U.S. (SUS) | | | | | | |
| Name | Category | Wind speed (km/h) | Landfall | Name | Category | Wind speed (km/h) | Landfall |
| PALOMA | 2 | 118\.52 | 8 November 2008 | DELTA | 2 | 157\.42 | 9 October 2020 |
| RICHARD | 2 | 148\.16 | 25 October 2010 | GUSTAV | 2 | 171\.31 | 1 September 2008 |
| GRACE | 3 | 185\.2 | 19 August 2021 | IKE | 2 | 162\.97 | 7 September 2008 |
| IAN | 3 | 197\.23 | 27 September 2022 | ISABEL | 2 | 148\.16 | 18 September 2003 |
| SANDY | 3 | 185\.2 | 24 October 2012 | SALLY | 2 | 166\.68 | 16 September 2020 |
| DENNIS | 4 | 213\.9 | 8 July 2005 | DENNIS | 3 | 175\.94 | 10 July 2005 |
| MARIA \* | 4 | 212\.98 | 20 September 2017 | IVAN | 3 | 180\.57 | 16 September 2004 |
| MICHELLE | 4 | 203\.72 | 4 November 2001 | KATRINA | 3 | 194\.46 | 25 August 2005 |
| WILMA | 4 | 185\.2 | 22 October 2005 | RITA | 3 | 175\.01 | 24 September 2005 |
| GUSTAV | 4 | 236\.13 | 30 August 2008 | ZETA | 3 | 185\.2 | 27 October 2020 |
| | | | | IDA | 4 | 212\.98 | 29 August 2021 |
| | | | | LAURA \* | 4 | 219\.46 | 27 August 2020 |
| | | | | MICHAEL | 5 | 259\.28 | 10 October 2018 |
| South Gulf (SG) | West Gulf (WG) | | | | | | |
| Name | Category | Wind speed (km/h) | Landfall | Name | Category | Wind speed (km/h) | Landfall |
| DELTA | 2 | 166\.68 | 7 October 2020 | ALEX | 2 | 166\.68 | 1 July 2010 |
| ERNESTO | 2 | 138\.9 | 8 August 2012 | EMILY | 3 | 185\.2 | 20 July 2005 |
| ISIDORE | 3 | 185\.2 | 22 September 2002 | HARVEY \* | 4 | 194\.46 | 26 August 2017 |
| KARL | 3 | 175\.94 | 17 September 2010 | | | | |
| EMILY \* | 4 | 212\.98 | 18 July 2005 | | | | |
| DEAN | 5 | 235\.2 | 21 August 2007 | | | | |
![]()
**Table 2.** Mean percent change in EVI AUC by subregion.
**Table 2.** Mean percent change in EVI AUC by subregion.
| Subregion | 1st Winter | 1st Year |
|---|---|---|
| Florida Peninsula (FP) | −0.02 | −1.25 |
| Central America (CA) | −9.69 | 0\.3 |
| Caribbean (C) | 0\.39 | −1.28 |
| Southeast U.S. (SUS) | −2.58 | −2.78 |
| South Gulf (SG) | 2\.33 | 0\.1 |
| West Gulf (WG) | −0.72 | −9.23 |
![]()
**Table 3.** Immediate and secondary response of EVI AUC for forest and all lands for the six hurricane case studies. Units are percent change in the sum of the 16-day EVI values under the winter or annual LSP curve.
**Table 3.** Immediate and secondary response of EVI AUC for forest and all lands for the six hurricane case studies. Units are percent change in the sum of the 16-day EVI values under the winter or annual LSP curve.
| Cases | All Lands | Forest | | |
|---|---|---|---|---|
| 1st Winter | 1st Year | 1st Winter | 1st Year | |
| Charley (FP) | −2.32 | 2\.74 | −2.47 | 1\.17 |
| Laura (SUS) | −7.88 | −9.43 | −10.23 | −13.64 |
| Harvey (WG) | −9.47 | −9.83 | −2.25 | −4.72 |
| Iota (CA) | −23.53 | −5.17 | −35.84 | −2.96 |
| Emily (SG) | 6\.44 | −6.63 | 11\.4 | −4.9 |
| Maria (C) | −7.85 | 0\.9 | −1.12 | 1\.64 |
| | |
|---|---|
| | **Disclaimer/Publisher’s Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the [Creative Commons Attribution (CC BY) license](https://creativecommons.org/licenses/by/4.0/).
## Share and Cite
**MDPI and ACS Style**
Topete-Pozas, C.; Norman, S.P. Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022). *Forests* **2026**, *17*, 419. https://doi.org/10.3390/f17040419
**AMA Style**
Topete-Pozas C, Norman SP. Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022). *Forests*. 2026; 17(4):419. https://doi.org/10.3390/f17040419
**Chicago/Turabian Style**
Topete-Pozas, Carlos, and Steven P. Norman. 2026. "Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022)" *Forests* 17, no. 4: 419. https://doi.org/10.3390/f17040419
**APA Style**
Topete-Pozas, C., & Norman, S. P. (2026). Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022). *Forests*, *17*(4), 419. https://doi.org/10.3390/f17040419
Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details [here](https://www.mdpi.com/about/announcements/784).
## Article Metrics
No
No
### Article Access Statistics
For more information on the journal statistics, click [here](https://www.mdpi.com/journal/forests/stats).
Multiple requests from the same IP address are counted as one view.
[Zoom]() \| [Orient]() \| [As Lines]() \| [As Sticks]() \| [As Cartoon]() \| [As Surface]() \| [Previous Scene]() \| [Next Scene]()
## Cite
Export citation file: [BibTeX]() \| [EndNote]() \| [RIS]()
**MDPI and ACS Style**
Topete-Pozas, C.; Norman, S.P. Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022). *Forests* **2026**, *17*, 419. https://doi.org/10.3390/f17040419
**AMA Style**
Topete-Pozas C, Norman SP. Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022). *Forests*. 2026; 17(4):419. https://doi.org/10.3390/f17040419
**Chicago/Turabian Style**
Topete-Pozas, Carlos, and Steven P. Norman. 2026. "Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022)" *Forests* 17, no. 4: 419. https://doi.org/10.3390/f17040419
**APA Style**
Topete-Pozas, C., & Norman, S. P. (2026). Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022). *Forests*, *17*(4), 419. https://doi.org/10.3390/f17040419
Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details [here](https://www.mdpi.com/about/announcements/784).
[*clear*]()
*[Forests](https://www.mdpi.com/journal/forests)*, EISSN 1999-4907, Published by MDPI
[RSS](https://www.mdpi.com/rss/journal/forests) [Content Alert](https://www.mdpi.com/journal/forests/toc-alert)
### Further Information
[Article Processing Charges](https://www.mdpi.com/apc) [Pay an Invoice](https://www.mdpi.com/about/payment) [Open Access Policy](https://www.mdpi.com/openaccess) [Contact MDPI](https://www.mdpi.com/about/contact) [Jobs at MDPI](https://careers.mdpi.com/)
### Guidelines
[For Authors](https://www.mdpi.com/authors) [For Reviewers](https://www.mdpi.com/reviewers) [For Editors](https://www.mdpi.com/editors) [For Librarians](https://www.mdpi.com/librarians) [For Publishers](https://www.mdpi.com/publishing_services) [For Societies](https://www.mdpi.com/societies) [For Conference Organizers](https://www.mdpi.com/conference_organizers)
### MDPI Initiatives
[Sciforum](https://sciforum.net/) [MDPI Books](https://www.mdpi.com/books) [Preprints.org](https://www.preprints.org/) [Scilit](https://www.scilit.com/) [SciProfiles](https://sciprofiles.com/?utm_source=mpdi.com&utm_medium=bottom_menu&utm_campaign=initiative) [Encyclopedia](https://encyclopedia.pub/) [JAMS](https://jams.pub/) [Proceedings Series](https://www.mdpi.com/about/proceedings)
### Follow MDPI
[LinkedIn](https://www.linkedin.com/company/mdpi) [Facebook](https://www.facebook.com/MDPIOpenAccessPublishing) [X](https://x.com/MDPIOpenAccess)
© 1996-2026 MDPI (Basel, Switzerland) unless otherwise stated
[Disclaimer]()
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
[Terms and Conditions](https://www.mdpi.com/about/terms-and-conditions) [Privacy Policy](https://www.mdpi.com/about/privacy)
We use cookies on our website to ensure you get the best experience.
Read more about our cookies [here](https://www.mdpi.com/about/privacy).
[Accept](https://www.mdpi.com/accept_cookies)
## Share Link
[Copy]()
[*clear*]()
## Share
https://www.mdpi.com/3815358
[*clear*]()
[Back to TopTop](https://www.mdpi.com/1999-4907/17/4/419) |
| Readable Markdown | [](https://www.mdpi.com/journal/forests)
Article Menu
Font Type:
*Arial* *Georgia* *Verdana*
Font Size:
Aa Aa Aa
Line Spacing:
** ** **
Column Width:
** ** **
Background:
Open AccessFeature PaperArticle
by Carlos Topete-Pozas [](https://orcid.org/0000-0003-0240-2823) andSteven P. Norman \*[](https://orcid.org/0000-0003-2080-9774)
US Department of Agriculture Forest Service, Southern Research Station, Eastern Forest Environmental Threat Assessment Center, 200 WT Weaver Blvd, Asheville, NC 28804, USA
\*
Author to whom correspondence should be addressed.
Submission received: 24 February 2026 / Revised: 17 March 2026 / Accepted: 25 March 2026 / Published: 27 March 2026
## Abstract
Hurricanes routinely damage forests across the North Atlantic Basin, yet efforts to characterize their impacts have had mixed subregional success. To elucidate these challenges, this study analyzed pre- and post-hurricane land surface phenology (LSP) for 44 moderate and strong hurricanes over 22 years using the Enhanced Vegetation Index (EVI). We statistically grouped storms based on their long-term climate attributes, then compared subregional impacts with wind speed and land cover. After accounting for wind speed, responses differed among the six subregions. The Southeast U.S. showed declines in EVI for the first winter and first year post storm, but this response was weak or absent elsewhere. The Central America region declined in the first winter but not in the subsequent growing season, while four other regions showed no increased impact with wind speed in either season. We then examined six category 4 hurricanes using a forest mask. In dry areas, drought-sensitive vegetation explained weak responses, whereas in the humid tropics, rapid refoliation or sprouting was common. These factors complicate optical remote sensing assessments. Rapid evaluations can mistake defoliation for more substantial damage, and delayed assessments can confuse EVI recovery with structural recovery. Results underscore the need for ecologically tailored monitoring approaches.
## 1\. Introduction
The North Atlantic Basin (NAB) commonly experiences extreme hurricanes with severe consequences for the natural and human systems of the region \[[1](https://www.mdpi.com/1999-4907/17/4/419#B1-forests-17-00419),[2](https://www.mdpi.com/1999-4907/17/4/419#B2-forests-17-00419),[3](https://www.mdpi.com/1999-4907/17/4/419#B3-forests-17-00419),[4](https://www.mdpi.com/1999-4907/17/4/419#B4-forests-17-00419)\]. Notably, the intensity and frequency of hurricanes in this region have been changing over the recent decades \[[3](https://www.mdpi.com/1999-4907/17/4/419#B3-forests-17-00419),[5](https://www.mdpi.com/1999-4907/17/4/419#B5-forests-17-00419),[6](https://www.mdpi.com/1999-4907/17/4/419#B6-forests-17-00419)\]. Between 1989 and 2018, an estimated one million km2 of forest and other land cover types were affected by hurricanes in the NAB, representing one of the highest damaged areas globally \[[7](https://www.mdpi.com/1999-4907/17/4/419#B7-forests-17-00419)\]. Indeed, the majority of studies regarding hurricane impacts on forest (67%) were conducted in this region \[[7](https://www.mdpi.com/1999-4907/17/4/419#B7-forests-17-00419)\].
Remote sensing has become a critical tool for assessing hurricane impacts and subsequent vegetation recovery at different spatio-temporal scales \[[8](https://www.mdpi.com/1999-4907/17/4/419#B8-forests-17-00419),[9](https://www.mdpi.com/1999-4907/17/4/419#B9-forests-17-00419)\]. Its usefulness for monitoring within the NAB is influenced by complex patterns of land cover and seasonal vegetation dynamics, both of which are informed by land surface phenology (LSP). Expanded use of LSP has improved how we understand disturbance impacts in light of seasonal and multi-year environmental and anthropogenic influences \[[10](https://www.mdpi.com/1999-4907/17/4/419#B10-forests-17-00419),[11](https://www.mdpi.com/1999-4907/17/4/419#B11-forests-17-00419),[12](https://www.mdpi.com/1999-4907/17/4/419#B12-forests-17-00419)\]. LSP-derived measures relate to important ecosystem processes, such as carbon and water cycles, timber markets, the fuels available for wildfire, and species interactions and connectivity \[[13](https://www.mdpi.com/1999-4907/17/4/419#B13-forests-17-00419)\]. LSP is particularly useful for tracking vegetation productivity, and this varies regionally and locally with vegetational composition, climate, soil, management practices, and weather disturbances, among other factors \[[14](https://www.mdpi.com/1999-4907/17/4/419#B14-forests-17-00419)\].
Recent research demonstrates the broad monitoring capabilities of LSP to characterize damage from hurricanes. Continuous satellite measurements give researchers the ability to track commonly used LSP parameters that capture key stages of seasonal vegetation behavior, such as the start of spring or fall or the peak of the growing season. In tropical or subtropical climates, hurricanes can impact the onset or progression of the late growing season contingent on when storms occur. This limitation makes LSP parameters that capture seasonal timing less reliable. Instead, for consistency, post-storm season integral phenometrics are more suitable, such as area under the curve (AUC) for all or a portion of the year. Such measures are recognized LSP phenometrics that provide a robust mechanism for understanding seasonal and annual variability \[[10](https://www.mdpi.com/1999-4907/17/4/419#B10-forests-17-00419),[15](https://www.mdpi.com/1999-4907/17/4/419#B15-forests-17-00419),[16](https://www.mdpi.com/1999-4907/17/4/419#B16-forests-17-00419),[17](https://www.mdpi.com/1999-4907/17/4/419#B17-forests-17-00419)\]. Along the US coast, studies have demonstrated the value of short- and long-term monitoring using different sensors and spectral indices including the 10 m resolution Sentinel 2 Normalized Difference Vegetation Index (NDVI) \[[18](https://www.mdpi.com/1999-4907/17/4/419#B18-forests-17-00419)\], 30 m resolution Landsat 5 and 7 and NDVI \[[19](https://www.mdpi.com/1999-4907/17/4/419#B19-forests-17-00419)\], and five vegetation indices from MODIS 1 km 16-day \[[20](https://www.mdpi.com/1999-4907/17/4/419#B20-forests-17-00419)\]. Others \[[21](https://www.mdpi.com/1999-4907/17/4/419#B21-forests-17-00419)\] assessed the effects of Hurricane Maria in Puerto Rico using Landsat 8 data with spectral mixture and statistical analyses in association with landform characteristics and forest structure.
Enhanced Vegetation Index (EVI) has been considered by several authors as a more suitable index to monitor forest disturbance in tropical regions because of EVI’s sensitivity to canopy variation in high biomass regions where NDVI saturates \[[22](https://www.mdpi.com/1999-4907/17/4/419#B22-forests-17-00419),[23](https://www.mdpi.com/1999-4907/17/4/419#B23-forests-17-00419),[24](https://www.mdpi.com/1999-4907/17/4/419#B24-forests-17-00419)\]. Research has demonstrated the effectiveness of MODIS EVI to assess the extent of immediate impacts from Hurricane Dean in the tropical forest of the Yucatan Peninsula, Mexico \[[25](https://www.mdpi.com/1999-4907/17/4/419#B25-forests-17-00419)\], and in Central America after Hurricane Felix \[[26](https://www.mdpi.com/1999-4907/17/4/419#B26-forests-17-00419)\]. Others \[[27](https://www.mdpi.com/1999-4907/17/4/419#B27-forests-17-00419)\] used MODIS EVI to assess hurricane damage and recovery in the northern Gulf and found that impacts vary by event, year, vegetation characteristics, and climate attributes \[[28](https://www.mdpi.com/1999-4907/17/4/419#B28-forests-17-00419)\], and monitored forest recovery along the northern Gulf comparing MODIS EVI with other spectral indices. In the Caribbean, ref. \[[29](https://www.mdpi.com/1999-4907/17/4/419#B29-forests-17-00419)\] used a MODIS disturbance index to capture both drought and hurricane impacts over four large Caribbean islands, and stressed the importance of considering land cover. In Cuba, ref. \[[30](https://www.mdpi.com/1999-4907/17/4/419#B30-forests-17-00419)\] assessed the extent and severity of damage to mangroves after Hurricane Irma with integrated use of MODIS and Sentinel-2 data. MODIS has been useful for addressing hurricane impacts in the tropics because its high temporal frequency helps overcome issues with cloud cover while isolating immediate responses and longer-term implications.
The NAB’s complex meteorological and land cover characteristics present challenges for monitoring hurricane impacts as methods may not be uniformly effective \[[31](https://www.mdpi.com/1999-4907/17/4/419#B31-forests-17-00419),[32](https://www.mdpi.com/1999-4907/17/4/419#B32-forests-17-00419)\]. While rapid optical remote sensing assessments are broadly effective for recognizing the footprint of potential impacts, commonly used techniques struggle with ecological precision, such as lasting structural damage and successional impacts in the tropics. Assessments in the humid tropics are especially time-sensitive with a high risk of misinterpreting recovery of the remote sensing index as something more. The most intensive research has taken place in the southern U.S., and this regional bias limits our understanding of monitoring effectiveness across the broader region \[[7](https://www.mdpi.com/1999-4907/17/4/419#B7-forests-17-00419)\]. The broader NAB region includes areas with substantially different vegetation and climate variability that may require different approaches for hurricane impact assessments to be accurate. Our research addresses these regional differences in the effectiveness of LSP monitoring after hurricanes to clarify important limitations and research needs. Our specific objectives include evaluating the importance of wind speed, land cover, and climate through the use of regional assessment and case studies.
## 2\. Materials and Methods
### 2\.1. Study Area and Hurricanes
We focused on major storms rather than tropical storms and Saffir–Simpson category 1 hurricanes to increase the likelihood of capturing wind damage to forests. We selected 44 category 2–5 hurricanes that made landfall within the NAB between the years 2001 to 2022 across the Southeast U.S., Mexico, the Caribbean, and Central America. We determined the landfall site and inland path for each hurricane using the International Best Track Archive for Climate Stewardship (IBTrACS) dataset \[[33](https://www.mdpi.com/1999-4907/17/4/419#B33-forests-17-00419)\]. For each hurricane, we defined a semi-uniform focal area by drawing a polygon that extended 25 km on each side of the track and 100 km inland. This standardized footprint is intended to emphasize the near-track region most likely to experience inner-core winds and associated severe wind damages and provides a consistent spatial basis for comparable impact estimation across storms and regions. In some cases, such as in the Caribbean, the polygons were somewhat smaller due to the size of the island or the need to exclude water bodies. We used each footprint to derive MODIS EVI data, land cover, and climate attributes. [Figure 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f001) shows an overview of the data used and workflow.
### 2\.2. Climate Regionalization
Various researchers have classified the NAB-related tropical cyclones in subregions according to ecology or climate factors \[[34](https://www.mdpi.com/1999-4907/17/4/419#B34-forests-17-00419),[35](https://www.mdpi.com/1999-4907/17/4/419#B35-forests-17-00419)\]. Our need was to characterize the basic climate drivers that influenced vegetation and recovery response, so instead of relying on these subdivisions, we derived gridded climate variables for each focal area directly. For each hurricane focal area, we derived the median monthly precipitation, maximum temperature, and vapor pressure deficit from 2000 to 2022 from the Climate Engine TerraClimate 4 km-Monthly dataset \[[36](https://www.mdpi.com/1999-4907/17/4/419#B36-forests-17-00419)\]. We then combined these climate time series for each focal area and analyzed their cross-seasonal and interannual similarities using non-hierarchical k\-means clustering \[[37](https://www.mdpi.com/1999-4907/17/4/419#B37-forests-17-00419)\] using R 2024.04.2 software \[[38](https://www.mdpi.com/1999-4907/17/4/419#B38-forests-17-00419)\]. We explored prescribing different numbers of clusters to balance the need for representation with geographic coherence. The final iteration assigned a statistically-derived membership for each of our 44 hurricanes to six distinct climate subregions.
### 2\.3. Hurricane Impacts and Their Drivers
High-frequency, multi-year LSP data broadly captures recurring seasonal behavior, hurricane impacts, and recovery. For each of the 44 focal areas, we calculated the areal mean 16-day EVI for a 22 year period (2001–2022) using 500 m MODIS Terra (MOD13A1) and Aqua (MYD13A1) data derived from Climate Engine (visited on July 2023). Instances of missing data from the source were handled using a non-parametric imputation method implemented in the missForest package in R 2024.04.2 software \[[38](https://www.mdpi.com/1999-4907/17/4/419#B38-forests-17-00419),[39](https://www.mdpi.com/1999-4907/17/4/419#B39-forests-17-00419)\]. This algorithm employs Random Forests to estimate and replace missing values in a dataset. To document the normal seasonal LSP behavior for each cluster across years, we combined the 22 years of 16-day EVI values for the member hurricanes for an annual curve showing the mean and standard deviation. We then compared the dates of the hurricanes for each cluster with their annual LSP to clarify the implications of LSP seasonality for monitoring.
To understand impacts for each focal area, we used the MODIS Terra & Aqua 500 m 16-day dataset. We clipped the EVI time series to have two full years before and two years after each storm’s calendar year and derived two seasonal measures: the area under the curve (AUC) for the first winter (December through February) and the AUC for the first full year after the storm. Impacts were quantified as the percent change in EVI AUC (dEVI) from the mean of the two pre-storm seasonal values. For each subregional cluster, we used regression to compare measured impacts to the respective hurricane’s median wind speed that was derived from IBTrACS \[[33](https://www.mdpi.com/1999-4907/17/4/419#B33-forests-17-00419)\]. We then performed Mann–Whitney (Wilcoxon) tests on the regressions to compare differences among regions for the two seasonal measures.
Land cover varied among the focal areas and this could potentially influence the vulnerability of vegetation to the storms and our interpretation of short and long-term impacts. These patterns and differences among focal areas were assessed in Google Earth Engine using the MCD12Q1.061 MODIS Land Cover Type Yearly Global 500 m dataset. We summarized the percent land cover for each subregional cluster. To simplify categories across the region we reclassified the original classes into seven broader groups: mixed forest and woody savannas were merged into “mixed forest”; urban, water, wetland and other were grouped as “other”; and croplands, savannas, evergreen broadleaf forest and deciduous broadleaf forest were used as defined in the original dataset.
### 2\.4. Local Case Studies
To achieve a more precise understanding of impacts for the six subregions, we analyzed the LSP on one representative category 4 hurricane for each subregional cluster. Hurricane Charley made landfall on 13 August 2004 near Cayo Costa, Florida, with a sustained wind of 210–249 km/h and caused an estimated \$14 billion in economic losses. Hurricane Laura made landfall near Cameron, Louisiana on 27 August 2020 causing an estimated of \$19 billion in economic losses. Hurricane Harvey, that was considered the second costliest U.S. tropical storm after Katrina, struck Texas on 26 August 2017 and caused an estimated damage of 125 billion. Hurricane Iota made landfall in Nicaragua on 17 November 2020, just two weeks after Eta passed over this same area. Hurricane Emily made landfall in the Yucatan Peninsula with sustained winds of 212 km/h. Hurricane Maria struck Puerto Rico on 20 September 2017 and resulted in severe damage. Detailed information on hurricane history and damage costs can be found at (<https://www.weather.gov/publications/assessments>, accessed on 24 February 2026, <https://www.noaa.gov/>, accessed on 24 February 2026).
For each of these six hurricanes, we derived a median EVI time series using the MODIS Terra (MOD13Q1) and Aqua (MYD13Q1) 250 m 16-day dataset and, respectively, that was compiled into an 8-day time series for all lands and forests alone. Our forest cover mask used the Global Land Cover ESA WorldCover 10 m v100 map (2020) dataset and was coarsened to 250 m. We smoothed each time series using a Savitsky–Golay algorithm in R 2024.04.2 software using a window of 23 observations to reduce noise \[[38](https://www.mdpi.com/1999-4907/17/4/419#B38-forests-17-00419),[40](https://www.mdpi.com/1999-4907/17/4/419#B40-forests-17-00419),[41](https://www.mdpi.com/1999-4907/17/4/419#B41-forests-17-00419),[42](https://www.mdpi.com/1999-4907/17/4/419#B42-forests-17-00419)\]. Masking allowed us to resolve the potential influence of land cover on LSP for these six case studies and to help interpret results from the prior analysis of the 44 hurricanes. We calculated similar measures of impact as we had for the prior analysis. Percent change in EVI AUC was calculated as the difference between the mean of the two years prior to the hurricane compared to the year after the event. The EVI AUC of the first winter after the storm (December to February) was compared the mean of the prior two winters to minimize the effects of interannual variability.
## 3\. Results
### 3\.1. Climate Regionalization
Clustering grouped the 44 hurricanes into six types defined by climate, which we labeled according to their primary subregional location ([Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002), [Table 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t001)). The Southeast U.S. (SUS) type had the most hurricanes at 13 and with all categories and a variety of wind speeds ranging from 148 to 259 km/h. The Caribbean (C) group included 10 hurricanes ranging from category 2 to 4. The Florida Peninsula (FP), South Gulf (SG) and Central America (CA), had six hurricanes each. The FP showed hurricanes categories 2 to 4 and the latter two from 2 to 5. With only three hurricanes represented, the West Gulf group had the fewest hurricanes, and these ranged from category 2 to 4. Clustering yielded strong spatial aggregations for FP, SUS, WG and CA. The location of C and SG were less well defined spatially, with the C group extending from Puerto Rico and Cuba to Belize and Mexico.
Land cover differed among the six subregions with respect to woody (deciduous, mixed, and evergreen forests) and other types ([Figure 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f003)). SG and CA averaged the most forest, with SUS and C averaging around 50% forest and FP and WG being predominantly savanna and grassland.
Subregions with the most forested land cover (SG and CA) show the highest median EVI while that with the least woody vegetation (WG) has the lowest EVI ([Figure 4](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f004)a). These differences in EVI are also consistent with the climate attributes of the subregions. CA’s high precipitation, warm temperature, and low VPD reflect its low latitude climate ([Figure 4](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f004)b–d). This contrasts with the least productive subregion, WG, that has the lowest precipitation and highest VPD ([Figure 4](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f004)b,d). SUS shows a broad range of EVI values and strong variation in temperature that reflects the strong seasonality of the higher-latitude climate and the large number of hurricanes included.
For all six subregions, hurricanes occurred near or following the peak of the growing season when vegetation was in normal seasonal decline ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)). The median and standard deviation for each subregion shows the interannual variation over the 21-year period for the member hurricanes. Tropical CA had the highest year-round median EVI that consistently exceeded 0.5. SG had considerably more seasonal variability despite peaking with a higher EVI in July at the onset of its hurricane season. C shows a low median EVI below 0.4 early in the year, with the highest values around 0.55 during summer, indicating relatively moderate productivity with pronounced seasonality. Of all the subregions, C has the broadest range of hurricane dates represented, which is consistent with the Caribbean’s geographic exposure. FP and SUS show lower peak EVIs with strong seasonal variation, and WG exhibits the lowest median EVI values of all six subregions with a weak growing season peak that is comparatively stable between May and October when the three hurricanes occurred.
### 3\.2. Hurricane Impacts
Of the six subregions, only SUS, FP and WG showed declines in EVI for both the first winter and first growing season after the storm, and the decline increased across those two periods ([Table 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t002)). While higher than expected for winter, C also shows a potential hurricane-associated decline for the first year after the storm. In contrast, more tropical CA showed the strongest mean decline of any subregion for the first winter, but none for the subsequent growing season. SG showed no measured decline on average, as EVI increased for both seasonal measures. Of all the subregions, SUS’s decline was most consistent across seasons.
Wind speed and EVI decline show an inconsistent relationship across subregions. For the first winter after the storms, stronger winds were associated with more decline for SUS (p = 0) and CA (p = 0.005), but other subregions showed no anticipated decline or pattern ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)a). For the first year after the storm, SUS consistently showed a decline in impacts with increased wind speed (p = 0), but no other subregion showed this relationship ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)b).
### 3\.3. Local Case Studies
The six selected category 4 hurricanes show generally consistent results for all lands compared to their respective subregions ([Table 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t002) and [Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003)). Five of the six hurricanes declined for the first winter, with Emily (SG) being the exception. Hurricane Maria (C) differed from its subregion’s response by also showing a first winter decline. For the first year after the storm, neither Maria (C) nor Charley (FP) showed a decline, but Iota (CA) did, although much less than in winter and unlike the response of the CA subregion overall. From the first winter to the first year, half of the cases show further decline while half improved, with Iota (CA) showing the most remarkable improvement.
The isolated forest responses of these six cases are directionally similar to their corresponding all-land responses but often with very different magnitudes ([Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003)). All cases but Emily (SG) showed an initial drop in forest EVI, which was a directionally consistent but stronger increase than its all-lands response. Note that Emily occurred in mid-July, which was one to four months earlier than the dates of the other five cases ([Table 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t001)), and has extensive broadleaf forest cover ([Figure 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f003)). Iota (CA)’s forest decline was much stronger than its all-lands decline, but its EVI first-year recovery was greater. Maria (C) and Charley (FP) appear to have also recovered quickly. The forest EVI behavior of Iota, Maria and Charley is the opposite of what occurred with Laura (SUS), Harvey (WG) and Emily (SG); each of these showed increased EVI departure from the first winter to the first year after the storm.
Comparison of the five-year LSP behavior for forests and all lands shows a broadly similar form prior to the storms with Iota (CA) being most different ([Figure 7](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f007)). Hurricane impacts are less than the seasonal variation in EVI. There are often recurring differences in cover types in terms of their seasonal amplitude or peak timing. Note in particular the earlier forest EVI peak for Laura (SUS), Harvey (WG), and Emily (SG).
As LSP measures can vary among years for reasons other than disturbance such as seasonal weather, the EVI profiles for the two pre-hurricane years shown in [Figure 7](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f007) can inform the interpretation and attribution of change shown in [Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003). Note the more irregular two-year pre-storm LSP variability for Harvey (WG), Iota (CA), Emily (SG), and Maria (C) for the first winter or full year. Also note the more complex, less predictable LSP behavior of both forest and all lands for Iota (CA) over all five years.
## 4\. Discussion
In this study, we investigated the impacts of 44 hurricanes that made landfall across six distinct subregions of the NAB using remote sensing. We show that seasonal insights from LSP help clarify and nuance our understanding of hurricane impacts, given the major climatic and ecological differences that are present across the region. Using two measures, the first winter after the storm and the first year after the storm, we demonstrated that higher wind speeds do not consistently correlate with greater storm impacts. For example, significant declines in EVI with increasing wind speed were observed only in the Southeast U.S., while other subregions showed weak or non-significant relationships ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)). Phenologically, when hurricanes occurred during the mid growing season, as indicated on [Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005), refoliation and rapid EVI recovery may have reduced our ability to recognize the magnitude of structural impacts. These ecological and technological limitations suggest the quality of remote sensing assessments using this common technique, particularly for tropical areas that have had the least field validations.
Isolating land cover types is a necessity. Non-woody vegetation types, such as grasslands or pasture, are biologically insensitive to wind damage, but they are often hypersensitive to unrelated climate variation that drives annual productivity. For analyses of change over time, the presence of these climate-sensitive types can distort the reliability of baseline conditions and the integrity of the change measure used to evaluate the post-storm vegetation state. This is a hazard to mixed land cover assessments when coarse-grid imagery is used, as non-targeted vegetation behavior can influence results. In heterogeneous landscapes, forest impacts may only be isolated through use of higher resolution imagery and masking that targets the susceptible vegetation. However, even when a high-resolution forest mask is available, it can be challenging to obtain high-resolution cloud-free imagery to conduct assessments at scale.
The phenological status of a forest at the time of the storm can affect the accuracy of impact assessments. An immediate response may indicate either structural tree damage or ephemeral leaf stripping, given the sensitivity of most spectral indices to foliage. Seasonally deciduous forest types may be particularly vulnerable to leaf stripping late in their growing season when their foliage is fragile due to senescence, and that is when most hurricanes occur ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)). If the goal is to map structural damage in a rapid assessment, not just the general footprint of the storm more generally, leaf stripping can give false positives—a hazard that can be further aggravated with change detection approaches when the timing of senescence varies from year to year. To avoid this, we conducted our first impact assessment during the first winter rather than immediately after the storm ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)), but this may have introduced a different problem as storms that occurred during the mid growing season may have experienced rapid refoliation prior to the winter assessment that persists. In the tropics, having a different time-since-storm could have affected the consistency of our first-winter measure among hurricanes. Remeasuring impacts across more than the two seasons used in this study may provide a way to isolate these complex phenological factors, but that can be difficult to achieve at scale in areas with frequent cloud cover. Very rapid assessments can also be misleading when hurricane rain provides drought relief through a short-lived resurgence of stressed vegetation, and this is a particularly confounding factor for areas with mixed land cover, pasture and grass. It is unclear if this rain response affected our results given the time of our winter assessment, but it could help explain surges of greenness reported by others using NDVI time series \[[43](https://www.mdpi.com/1999-4907/17/4/419#B43-forests-17-00419)\].
Crown and bole sprouting can initiate just weeks after a hurricane in tropical forests, and this phenomenon can obscure structural damage \[[44](https://www.mdpi.com/1999-4907/17/4/419#B44-forests-17-00419),[45](https://www.mdpi.com/1999-4907/17/4/419#B45-forests-17-00419)\]. This phenomenon is less relevant for portions of SUS where hurricane damage to the predominant industrial loblolly pine forests often leads to tree mortality \[[46](https://www.mdpi.com/1999-4907/17/4/419#B46-forests-17-00419)\]. However, even north of SUS in the Appalachians, hardwood sprouting was prevalent after 2024’s hurricane Helene during the first growing season after the storm (personal observations of the second author, July 2025), suggesting that this is a widespread wind adaptive response. Sprouting vegetation can reduce the accuracy of impact assessments that use optical remote sensing, and this limits our understanding of forest resistance and recovery. Used alone or in combination, alternative technologies that capture change in vegetation height, such as high-resolution aerial LiDAR, largely bypass much of this problem as sprouting can occur on crown-damaged, leaning and fallen trees. Unfortunately, precise structural assessments using this technology take time and the extent of the analysis can be restricted \[[47](https://www.mdpi.com/1999-4907/17/4/419#B47-forests-17-00419)\].
### Subregions
Our six subregions differ in terms of climate variability, land cover and vegetational sensitivity, and phenology. These attributes are ecologically inter-related and they provide useful ways to frame post-hurricane impact assessments. Awareness of these subregional attributes can help identify where research is lacking and help tailor post-hurricane monitoring to improve its effectiveness.
The Southeastern U.S. (SUS) stands out as a reliable area for long- and short-term monitoring of hurricane effects, and this is consistent with prior studies \[[7](https://www.mdpi.com/1999-4907/17/4/419#B7-forests-17-00419),[48](https://www.mdpi.com/1999-4907/17/4/419#B48-forests-17-00419),[49](https://www.mdpi.com/1999-4907/17/4/419#B49-forests-17-00419)\]. Impacts are resolvable during both winter and the subsequent growing season with impacts increasing at higher wind speeds, as expected ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)). Other research has shown that damage to the industrial pine of this subregion is especially well-captured, as is the post-hurricane management response, but impacts to deciduous and mixed forests are harder to resolve \[[18](https://www.mdpi.com/1999-4907/17/4/419#B18-forests-17-00419)\]. Isolating leaf stripping from structural damage for deciduous forests can be challenging for storms that occur during fall senescence, but strong winter seasonality limits the potential for rapid hardwood sprouting.
In contrast, monitoring hurricane impacts in Central America (CA) is more challenging, particularly for the year after the storm. This is apparently due to the warm-wet tropical climate and high ecosystem productivity that causes rapid EVI recovery for the evergreen broadleaf and mixed forest types. For the first winter after the storm, we documented EVI declines that strengthened with wind speed, but we found no consistent evidence of impacts thereafter. If extreme hurricanes cause lasting structural damage to these forests, it is likely being obscured by refoliation, because over 70% of the land cover of CA is forest. Category 4 Iota had the strongest initial departure of any of the 44 hurricanes considered, but despite category 5 winds, Felix had less impact, although its impact was directionally similar to Iota over the two periods ([Table 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t001), [Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)). This may be because Felix’s focal area had substantially more grassland which is less sensitive to damage than forest ([Appendix A](https://www.mdpi.com/1999-4907/17/4/419#app1-forests-17-00419), [Table A1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t0A1)). The irregular year-to-year LSP shown for Hurricane Iota ([Figure 7](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f007)d) complicates assessment, and this may be caused by the quality of the remote sensing data given the frequency of clouds in this humid region. In CA, hurricanes tend to occur in late November, which is phenologically later than that other subregions ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)). That difference in timing may affect the sensitivity of the deciduous canopy to defoliation or how aggressively hardwood refoliation occurs. Monitoring damage during first winter may be the most feasible for the CA subregion, but it is challenging to isolate structural impacts.
Only three of 44 hurricanes were included in the West Gulf (WG) subregion, and Harvey, the strongest, had the lowest mean wind speed of the six category 4 case studies ([Table 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t001)). Consistent with its warm-dry climate, WG had the lowest EVI across seasons and the least forest cover ([Figure 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f003), [Figure 4](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f004) and [Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005), [Appendix A](https://www.mdpi.com/1999-4907/17/4/419#app1-forests-17-00419), [Table A1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t0A1)). At less than 10% forest, the subregion’s vegetation should be inherently less sensitive to hurricane damage, given that grass is unlikely to retain damage into the next growing season, yet both forests and all lands show notable declines ([Table 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t002) and [Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003)). It is particularly surprising that the first-year departure of all land cover is roughly double that of the forest for Harvey ([Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003)). This suggests a cause from something other than the storm, such as the influence of interannual climate variability. Grass is particularly sensitive to climate variability and drought \[[50](https://www.mdpi.com/1999-4907/17/4/419#B50-forests-17-00419),[51](https://www.mdpi.com/1999-4907/17/4/419#B51-forests-17-00419)\]. Our measure of storm impact, namely, change from a prior baseline, is particularly vulnerable to seasonal climate variability because either the pre- or post-disturbance value could be compromised. This problem appears to explain the first-year response of 2010’s category 2 Alex that, contrary to expectations, showed a stronger decline than the more intense storms for this subregion. Unlike the baseline used to calculate change, 2011’s first-year EVI was likely low given a severe drought as indicated by the North American Drought Monitor (<https://droughtmonitor.unl.edu/NADM/Maps.aspx>, accessed on 24 February 2026). This inherent limitation of monitoring in the subregion can reduce confidence in the accuracy of impact assessments, including those that rely on coarse-resolution forest masks. Results are also more challenging to interpret in this subregion with so few storms represented. For this subregion, forest monitoring requires a high spatial resolution analysis and sensitivity to interannual climate variability.
The South Gulf (SG) subregion is characterized by highly productive, seasonal forests with more than 80% forest cover, the highest of the six subregions ([Figure 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f003)). Annual EVI is the second-highest among the subregions, peaking as the highest but with a strong winter decline that shows high variability ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)). Hurricanes in SG occurred from July, at peak EVI, through September, which is phenologically earlier than for those of other subregions. Rapid refoliation occurs during the growing season, and early storm dates may reduce the usefulness of first winter and first year measures ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)). For example, Emily, the subregion’s case study, was the earliest of the six cases, and the fourth-earliest of all 44 storms analyzed ([Table 1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t001)). This adaptation of tropical hardwoods may be more strongly expressed when defoliation occurs during the mid growing season, so impact assessment may be more accurate when conducted immediately after the hurricane occurs, although monitoring at this time can be difficult for low lying areas due to lingering floodwaters and the prevalence of non-structural defoliation.
The ten hurricanes analyzed for the Caribbean (C) subregion show varied responses for the first winter and first year after their storms with respect to wind speed. Some storms show a decline while others show an increase in EVI for these two periods. A recent study from Turner et al., \[[30](https://www.mdpi.com/1999-4907/17/4/419#B30-forests-17-00419)\], on the effects of hurricane Irma in Cuba found that while there was a significant decrease in EVI in mangroves and wetlands, there was a widespread increase in EVI elsewhere, including for the dry forest. In our analysis, Maria showed an EVI decline for the first winter after the storm, but an increase in EVI during the first year after the storm. This behavior is consistent with de Beurs et al., \[[29](https://www.mdpi.com/1999-4907/17/4/419#B29-forests-17-00419)\] who documented decline from Maria in Puerto Rico, with a quick recovery starting 8–12 weeks after the storm which they suggested might be associated with a growth surge from heavy rainfall. The timing of hurricanes here with respect to the region’s phenology suggests a strong potential for rapid refoliation and growth before the end of the growing season much of the time ([Figure 5](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f005)). In addition, with less forest cover than SG and CA and rugged topography that fragments forest cover, EVI’s sensitivity to post-storm weather variability may also be high. Similarly to SG, resolving hurricane impacts in this subregion can be challenging, but insights are possible soon after the storm. Use of high resolution imagery can help isolate the confounding responses of cover types.
Only about 10 percent of the Florida Peninsula (FP) subregion is forested, as savannas, grasslands, croplands, and other land cover types predominate. These latter types are less sensitive to damage, and this likely explains the non-responsiveness of EVI regardless of wind speed ([Figure 6](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f006)). The response of FP’s Hurricane Charley shows a short-term decline followed by rapid recovery for both all lands and forests, and this behavior is generally consistent with our detailed case studies for C and CA ([Table 3](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t003)). This may reflect the rapid sprouting of trees and shrubs that may or may not have substantial structural damage. While limited in distribution, hurricane damage to south Florida’s mangrove forests has received the most research attention and the homogeneity of the type makes it more easily monitored. The mangroves here have had four declines from severe hurricanes during the last three decades with slow recovery and a long-term decline \[[52](https://www.mdpi.com/1999-4907/17/4/419#B52-forests-17-00419)\]. Elsewhere, successful hurricane monitoring in the FP will likely benefit from high-resolution analyses that isolate the patchy woody plant cover.
## 5\. Conclusions
This study identified inconsistencies in the predicted relationship between wind speed and hurricane damage. Subtropical SUS showed increased damage with wind speed—a pattern that remained across seasons. This anticipated relationship was absent elsewhere, as EVI change often increased after severe disturbance. Confounding drivers include subregional differences in land cover and climate-related productivity. Reduced evidence of storm impacts by the first growing season was common the tropics, which is consistent with higher productivity and rapid vegetative recovery, but some tropical hurricane footprints showed increased departure over time. This suggests that our measures of change may capture a combination of disturbance impacts and background variation in climate or LSP that could affect our ability to recognize change.
Our results underscore the need for a strategic approach to monitoring post-hurricane impacts that is tailored to the unique ecological conditions and land-use configurations of each subregion. Our findings emphasize the value of early monitoring for the tropics where rapid sprouting can obscure structural damage. However, rapid assessments that occur during the same growing season as the storm can misrepresent minor defoliation as structural damage. Rapid refoliation can occur in areas with or without structural damage, thus confounding estimates of vegetation structural recovery. This is not to say that rapid assessments with coarse spatial resolution lack value, as they can suggest the footprint of potential impacts where more focused evaluations can be pursued. This research shows that by the first year after the storm, structural impacts are difficult to resolve in the tropics, which means that subsequent monitoring of structural recovery will be even more difficult. This affects our ability to accurately assess the dynamics and resilience of these disturbance-prone ecosystems using this technology.
Our research shows that optical techniques are more effective for assessing immediate and long-term impacts to temperate regions such as SUS. This is not just because more research has been conducted there, as analyses benefit from the subregion’s relatively predictable phenological behavior and the sensitivity of industrial pines to storm damage \[[18](https://www.mdpi.com/1999-4907/17/4/419#B18-forests-17-00419)\]. The strong seasonality of temperate latitude forests means that rapid sprouting and refoliation are less likely to be concerns here. Outside this subregion, impact and monitoring assessments are more likely to struggle with distinguishing the recovery shown by spectral indices from structural recovery. The latter, more elusive phenomenon is needed to characterize long-term impacts, ecosystem resilience, and the global implications of these frequent, large-scale storms \[[53](https://www.mdpi.com/1999-4907/17/4/419#B53-forests-17-00419)\].
Hurricane impact monitoring will benefit from understanding how and why vegetation types vary from year to year normally. Non-woody vegetation may exhibit a surge in late-season growth from the influx of tropical storm rainfall in ways that interfere with rapid forest damage assessment, but it is unclear if this short-term mechanism affects the rate of woody plant refoliation. The cross-seasonal and year-to-year productivity of non-woody types can also vary in response to drought. Thus, weather variability can affect the integrity of either the normal baseline condition or the post-disturbance indicator of damage. Where vegetation is fragmented or mixed, monitoring at high spatial resolution with forest masks can reduce this problem. Conversely, hurricane footprints with relatively homogeneous forest cover have less issues with mixed responses from fractional cover and are more amenable to assessment from coarse resolution imagery.
Better use of LSP can improve our understanding of hurricane impacts across the ecologically and climatically diverse NAB. LSP encapsulates vegetation type and productivity, the seasonal and interannual dynamics that need consideration, and the actual storm impacts. LSP is more difficult to leverage in the humid tropics where high-resolution analyses, synthetic aperture radar (SAR) and (LiDAR) technologies that resolve structural change and recovery more precisely appear to be warranted.
## Author Contributions
Conceptualization and methodology, S.P.N. and C.T.-P.; analysis and investigation, C.T.-P. and S.P.N.; data curation, C.T.-P. and S.P.N.; writing—original draft preparation, C.T.-P. and S.P.N.; writing—review and editing C.T.-P. and S.P.N.; visualization, C.T.-P. and S.P.N.; project administration: S.P.N. All authors have read and agreed to the published version of the manuscript.
## Funding
This research received no external funding.
## Data Availability Statement
The data supporting the conclusions of this article will be made available by the authors upon request.
## Acknowledgments
This research was supported in part by an appointment to the United States Forest Service (USFS) Research Participation Program administered by the Oak Ridge Institute for Science and Education (ORISE) through an interagency agreement between the U.S. Department of Energy (DOE) and the U.S. Department of Agriculture (USDA). The findings and conclusions in this publication are those of the authors and should not be construed to represent any official USDA or U.S. Government determination or policy.
## Conflicts of Interest
The authors declare that they have no commercial or associative interests that represent conflicts of interest in connection with the article submitted.
## Abbreviations
The following abbreviations are used in this manuscript:
| | |
|---|---|
| LSP | Land Surface Phenology |
| EVI | Enhanced Vegetation Index |
| AUC | Area Under the Curve |
| NAB | North Atlantic Basin |
| C | Caribbean |
| CA | Central America |
| FP | Florida Peninsula |
| SG | South Gulf |
| SUS | Southeast United States |
| WG | West Gulf |
## Appendix A
**Table A1.** Percent of land cover types within the 44 hurricanes focal areas derived from MODIS.
**Table A1.** Percent of land cover types within the 44 hurricanes focal areas derived from MODIS.
| | | | | | | | |
|---|---|---|---|---|---|---|---|
| Florida Peninsula (FP) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| FRANCES | 1\.1 | 0\.4 | 0\.4 | 6\.7 | 1\.2 | 8 | 82\.2 |
| IRMA | 0\.4 | 0\.7 | 1\.2 | 7\.6 | 12\.5 | 23\.3 | 54\.3 |
| JEANNE | 0\.8 | 0\.1 | 0\.4 | 7\.6 | 1\.7 | 11\.8 | 77\.6 |
| WILMA | 1\.9 | 0\.5 | 3\.2 | 3\.9 | 18\.7 | 34\.1 | 37\.8 |
| CHARLEY \* | 0\.2 | 0\.1 | 0\.7 | 12\.4 | 2\.9 | 9\.1 | 74\.6 |
| IAN | 0\.1 | 0\.1 | 0\.8 | 9\.9 | 3\.1 | 9 | 77\.1 |
| Caribbean (C) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| PALOMA | 31\.8 | 12 | 6\.4 | 1\.3 | 17\.9 | 1\.6 | 29 |
| RICHARD | 0\.5 | 0 | 72\.9 | 8\.8 | 12\.2 | 2\.8 | 2\.7 |
| GRACE | 8 | 0 | 5\.7 | 2\.6 | 50\.9 | 2 | 30\.9 |
| IAN | 22\.4 | 0\.1 | 17\.3 | 5\.7 | 21\.9 | 5\.4 | 27\.2 |
| SANDY | 37\.7 | 0\.3 | 18\.1 | 1\.4 | 24\.2 | 3\.6 | 14\.6 |
| DENNIS | 18\.4 | 1\.6 | 19\.6 | 1\.6 | 11\.9 | 16\.4 | 30\.5 |
| MARIA \* | 6\.9 | 0 | 25 | 7\.2 | 41\.8 | 14\.5 | 4\.6 |
| MICHELLE | 31\.6 | 7\.7 | 13\.1 | 0\.6 | 9\.5 | 6 | 31\.5 |
| WILMA | 0 | 0\.1 | 67\.3 | 3\.4 | 20\.3 | 8\.6 | 0\.3 |
| GUSTAV | 27\.9 | 0\.2 | 15\.6 | 3\.3 | 16\.5 | 5\.3 | 31\.3 |
| South Gulf (SG) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| DELTA | 0\.3 | 2 | 71\.9 | 2\.6 | 20\.4 | 1\.1 | 1\.7 |
| ERNESTO | 0\.1 | 0\.1 | 80\.5 | 2\.9 | 11\.7 | 3\.8 | 0\.9 |
| ISIDORE | 0\.3 | 32\.1 | 0 | 4\.6 | 53\.6 | 4 | 5\.5 |
| KARL | 43\.1 | 1\.8 | 14\.1 | 0\.9 | 14\.8 | 1\.8 | 23\.6 |
| EMILY \* | 0\.1 | 10\.9 | 78\.6 | 0\.6 | 9\.2 | 0\.5 | 0 |
| DEAN | 0 | 0\.2 | 81\.3 | 2\.4 | 11\.7 | 3\.7 | 0\.7 |
| Central America (CA) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| BETA | 0 | 0 | 29 | 1\.9 | 33\.5 | 3\.7 | 31\.8 |
| IRIS | 0 | 0 | 56\.4 | 6\.9 | 30\.4 | 0\.2 | 6 |
| OTTO | 0 | 0 | 52\.7 | 1\.4 | 25\.7 | 0\.9 | 19\.3 |
| ETA | 0 | 0 | 43\.6 | 9\.8 | 33\.5 | 8\.5 | 4\.6 |
| IOTA \* | 0 | 0 | 45\.2 | 4\.2 | 32\.7 | 6\.2 | 11\.7 |
| FELIX | 0 | 0 | 40\.7 | 18\.9 | 13 | 8\.1 | 19\.2 |
| Southeast U.S. (SUS) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| DELTA | 39\.4 | 3\.4 | 2\.4 | 4\.7 | 8\.7 | 15 | 26\.5 |
| GUSTAV | 2\.1 | 6 | 0\.1 | 1\.9 | 6\.2 | 75\.3 | 8\.4 |
| IKE | 0\.9 | 2\.6 | 8\.7 | 6\.4 | 39\.4 | 16\.4 | 25\.7 |
| ISABEL | 21\.3 | 6\.2 | 3\.1 | 1\.6 | 55\.9 | 9\.9 | 2\.1 |
| SALLY | 10\.8 | 3\.1 | 14\.4 | 7\.6 | 51\.1 | 3 | 10 |
| DENNIS | 5\.7 | 11\.3 | 10\.4 | 3\.8 | 62\.1 | 2\.6 | 4\.1 |
| IVAN | 4\.5 | 19\.4 | 12\.8 | 3 | 47\.4 | 7\.6 | 5\.3 |
| KATRINA | 0 | 11 | 6\.8 | 2\.2 | 66\.7 | 5\.5 | 7\.8 |
| RITA | 0\.2 | 5\.8 | 11\.5 | 2\.3 | 40\.6 | 17\.9 | 21\.7 |
| ZETA | 1\.3 | 2\.6 | 0 | 2\.7 | 4\.2 | 62\.6 | 26\.5 |
| IDA | 8\.6 | 15\.7 | 0 | 3 | 12 | 39 | 21\.7 |
| LAURA \* | 4\.1 | 2 | 13\.4 | 4\.4 | 32\.4 | 19\.1 | 24\.6 |
| MICHAEL | 4\.9 | 5\.5 | 22\.5 | 3\.6 | 52\.4 | 6 | 5\.1 |
| West Gulf (WG) | | | | | | | |
| Name | Croplands | Deciduous Broadleaf Forests | Evergreen Broadleaf Forests | Grasslands | Mixed Forests | Other | Savannas |
| ALEX | 0\.4 | 4\.7 | 2\.5 | 44\.5 | 6 | 5\.9 | 35\.9 |
| EMILY | 5\.1 | 0\.1 | 0 | 66\.3 | 2\.4 | 15\.7 | 10\.3 |
| HARVEY \* | 4\.4 | 0\.1 | 0 | 25\.3 | 0 | 4\.6 | 65\.7 |
\* The hurricanes examined as case studies.
## References
1. Cole, J.; Nowak, D.J.; Greenfield, E.J. Potential Hurricane Wind Risk to US Rural and Urban Forests. J. For. **2021**, 119, 393–406. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Potential+Hurricane+Wind+Risk+to+US+Rural+and+Urban+Forests&author=Cole,+J.&author=Nowak,+D.J.&author=Greenfield,+E.J.&publication_year=2021&journal=J.+For.&volume=119&pages=393%E2%80%93406&doi=10.1093/jofore/fvab018)\] \[[CrossRef](https://doi.org/10.1093/jofore/fvab018)\]
2. Mallin, M.A.; Corbett, C.A. How Hurricane Attributes Determine the Extent of Environmental Effects: Multiple Hurricanes and Different Coastal Systems. Estuaries Coasts **2006**, 29, 1046–1061. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=How+Hurricane+Attributes+Determine+the+Extent+of+Environmental+Effects:+Multiple+Hurricanes+and+Different+Coastal+Systems&author=Mallin,+M.A.&author=Corbett,+C.A.&publication_year=2006&journal=Estuaries+Coasts&volume=29&pages=1046%E2%80%931061&doi=10.1007/BF02798667)\] \[[CrossRef](https://doi.org/10.1007/BF02798667)\]
3. Reed, K.A.; Wehner, M.F.; Zarzycki, C.M. Attribution of 2020 Hurricane Season Extreme Rainfall to Human-Induced Climate Change. Nat. Commun. **2022**, 13, 1905. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Attribution+of+2020+Hurricane+Season+Extreme+Rainfall+to+Human-Induced+Climate+Change&author=Reed,+K.A.&author=Wehner,+M.F.&author=Zarzycki,+C.M.&publication_year=2022&journal=Nat.+Commun.&volume=13&pages=1905&doi=10.1038/s41467-022-29379-1&pmid=35414063)\] \[[CrossRef](https://doi.org/10.1038/s41467-022-29379-1)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/35414063)\]
4. Sharma, A.; Ojha, S.K.; Dimov, L.D.; Vogel, J.G.; Nowak, J. Long-Term Effects of Catastrophic Wind on Southern US Coastal Forests: Lessons from a Major Hurricane. PLoS ONE **2021**, 16, e0243362. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Long-Term+Effects+of+Catastrophic+Wind+on+Southern+US+Coastal+Forests:+Lessons+from+a+Major+Hurricane&author=Sharma,+A.&author=Ojha,+S.K.&author=Dimov,+L.D.&author=Vogel,+J.G.&author=Nowak,+J.&publication_year=2021&journal=PLoS+ONE&volume=16&pages=e0243362&doi=10.1371/journal.pone.0243362)\] \[[CrossRef](https://doi.org/10.1371/journal.pone.0243362)\]
5. Marsooli, R.; Lin, N.; Emanuel, K.; Feng, K. Climate Change Exacerbates Hurricane Flood Hazards along US Atlantic and Gulf Coasts in Spatially Varying Patterns. Nat. Commun. **2019**, 10, 3785. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Climate+Change+Exacerbates+Hurricane+Flood+Hazards+along+US+Atlantic+and+Gulf+Coasts+in+Spatially+Varying+Patterns&author=Marsooli,+R.&author=Lin,+N.&author=Emanuel,+K.&author=Feng,+K.&publication_year=2019&journal=Nat.+Commun.&volume=10&pages=3785&doi=10.1038/s41467-019-11755-z&pmid=31439853)\] \[[CrossRef](https://doi.org/10.1038/s41467-019-11755-z)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/31439853)\]
6. Vecchi, G.A.; Landsea, C.; Zhang, W.; Villarini, G.; Knutson, T. Changes in Atlantic Major Hurricane Frequency since the Late-19th Century. Nat. Commun. **2021**, 12, 4054. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Changes+in+Atlantic+Major+Hurricane+Frequency+since+the+Late-19th+Century&author=Vecchi,+G.A.&author=Landsea,+C.&author=Zhang,+W.&author=Villarini,+G.&author=Knutson,+T.&publication_year=2021&journal=Nat.+Commun.&volume=12&pages=4054&doi=10.1038/s41467-021-24268-5&pmid=34257285)\] \[[CrossRef](https://doi.org/10.1038/s41467-021-24268-5)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/34257285)\]
7. Lin, T.C.; Hogan, J.A.; Chang, C.T. Tropical Cyclone Ecology: A Scale-Link Perspective. Trends Ecol. Evol. **2020**, 35, 594–604. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Tropical+Cyclone+Ecology:+A+Scale-Link+Perspective&author=Lin,+T.C.&author=Hogan,+J.A.&author=Chang,+C.T.&publication_year=2020&journal=Trends+Ecol.+Evol.&volume=35&pages=594%E2%80%93604&doi=10.1016/j.tree.2020.02.012)\] \[[CrossRef](https://doi.org/10.1016/j.tree.2020.02.012)\]
8. Gao, W.; Larjavaara, M. Wind Disturbance in Forests: A Bibliometric Analysis and Systematic Review. For. Ecol. Manag. **2024**, 564, 122001. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Wind+Disturbance+in+Forests:+A+Bibliometric+Analysis+and+Systematic+Review&author=Gao,+W.&author=Larjavaara,+M.&publication_year=2024&journal=For.+Ecol.+Manag.&volume=564&pages=122001&doi=10.1016/j.foreco.2024.122001)\] \[[CrossRef](https://doi.org/10.1016/j.foreco.2024.122001)\]
9. Hernandez, J.O.; Maldia, L.S.J.; Park, B.B.; Hernandez, J.O.; Maldia, L.S.J.; Park, B.B. Research Trends and Methodological Approaches of the Impacts of Windstorms on Forests in Tropical, Subtropical, and Temperate Zones: Where Are We Now and How Should Research Move Forward? Plants **2020**, 9, 1709. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Research+Trends+and+Methodological+Approaches+of+the+Impacts+of+Windstorms+on+Forests+in+Tropical,+Subtropical,+and+Temperate+Zones:+Where+Are+We+Now+and+How+Should+Research+Move+Forward?&author=Hernandez,+J.O.&author=Maldia,+L.S.J.&author=Park,+B.B.&author=Hernandez,+J.O.&author=Maldia,+L.S.J.&author=Park,+B.B.&publication_year=2020&journal=Plants&volume=9&pages=1709&doi=10.3390/plants9121709&pmid=33291785)\] \[[CrossRef](https://doi.org/10.3390/plants9121709)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/33291785)\]
10. Henebry, G.M.; de Beurs, K.M. Remote Sensing of Land Surface Phenology: A Prospectus. In Phenology: An Integrative Environmental Science; Schwartz, M.D., Ed.; Springer: Dordrecht, The Netherlands, 2013; pp. 385–411. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Remote+Sensing+of+Land+Surface+Phenology:+A+Prospectus&author=Henebry,+G.M.&author=de+Beurs,+K.M.&publication_year=2013&pages=385%E2%80%93411)\]
11. Pomara, L.Y.; Lee, D.C.; Brooks, B.-G.; Hargrove, W.W. Using Land Surface Phenology and Information Theory to Assess and Map Complex Landscape Dynamics. Landsc. Ecol. **2024**, 39, 203. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Using+Land+Surface+Phenology+and+Information+Theory+to+Assess+and+Map+Complex+Landscape+Dynamics&author=Pomara,+L.Y.&author=Lee,+D.C.&author=Brooks,+B.-G.&author=Hargrove,+W.W.&publication_year=2024&journal=Landsc.+Ecol.&volume=39&pages=203&doi=10.1007/s10980-024-02005-9)\] \[[CrossRef](https://doi.org/10.1007/s10980-024-02005-9)\]
12. White, M.A.; de Beurs, K.M.; Didan, K.; Inouye, D.W.; Richardson, A.D.; Jensen, O.P.; O’Keefe, J.; Zhang, G.; Nemani, R.R.; van Leeuwen, W.J.D.; et al. Intercomparison, Interpretation, and Assessment of Spring Phenology in North America Estimated from Remote Sensing for 1982-2006. Glob. Change Biol. **2009**, 15, 2335–2359. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Intercomparison,+Interpretation,+and+Assessment+of+Spring+Phenology+in+North+America+Estimated+from+Remote+Sensing+for+1982-2006&author=White,+M.A.&author=de+Beurs,+K.M.&author=Didan,+K.&author=Inouye,+D.W.&author=Richardson,+A.D.&author=Jensen,+O.P.&author=O%E2%80%99Keefe,+J.&author=Zhang,+G.&author=Nemani,+R.R.&author=van+Leeuwen,+W.J.D.&author=et+al.&publication_year=2009&journal=Glob.+Change+Biol.&volume=15&pages=2335%E2%80%932359&doi=10.1111/j.1365-2486.2009.01910.x)\] \[[CrossRef](https://doi.org/10.1111/j.1365-2486.2009.01910.x)\]
13. Caparros-Santiago, J.A.; Rodriguez-Galiano, V.; Dash, J. Land Surface Phenology as Indicator of Global Terrestrial Ecosystem Dynamics: A Systematic Review. ISPRS J. Photogramm. Remote Sens. **2021**, 171, 330–347. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Land+Surface+Phenology+as+Indicator+of+Global+Terrestrial+Ecosystem+Dynamics:+A+Systematic+Review&author=Caparros-Santiago,+J.A.&author=Rodriguez-Galiano,+V.&author=Dash,+J.&publication_year=2021&journal=ISPRS+J.+Photogramm.+Remote+Sens.&volume=171&pages=330%E2%80%93347&doi=10.1016/j.isprsjprs.2020.11.019)\] \[[CrossRef](https://doi.org/10.1016/j.isprsjprs.2020.11.019)\]
14. Kooistra, L.; Berger, K.; Brede, B.; Graf, L.V.; Aasen, H.; Roujean, J.L.; Machwitz, M.; Schlerf, M.; Atzberger, C.; Prikaziuk, E.; et al. Reviews and Syntheses: Remotely Sensed Optical Time Series for Monitoring Vegetation Productivity. Biogeosciences **2024**, 21, 473–511. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Reviews+and+Syntheses:+Remotely+Sensed+Optical+Time+Series+for+Monitoring+Vegetation+Productivity&author=Kooistra,+L.&author=Berger,+K.&author=Brede,+B.&author=Graf,+L.V.&author=Aasen,+H.&author=Roujean,+J.L.&author=Machwitz,+M.&author=Schlerf,+M.&author=Atzberger,+C.&author=Prikaziuk,+E.&author=et+al.&publication_year=2024&journal=Biogeosciences&volume=21&pages=473%E2%80%93511&doi=10.5194/bg-21-473-2024)\] \[[CrossRef](https://doi.org/10.5194/bg-21-473-2024)\]
15. Cui, K.L.; Yang, J.L.; Dong, J.W.; Zhao, G.S.; Cui, Y.P. Comparing Different Spatial Resolutions and Indices for Retrieving Land Surface Phenology for Deciduous Broadleaf Forests. Remote Sens. **2023**, 15, 2266. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Comparing+Different+Spatial+Resolutions+and+Indices+for+Retrieving+Land+Surface+Phenology+for+Deciduous+Broadleaf+Forests&author=Cui,+K.L.&author=Yang,+J.L.&author=Dong,+J.W.&author=Zhao,+G.S.&author=Cui,+Y.P.&publication_year=2023&journal=Remote+Sens.&volume=15&pages=2266&doi=10.3390/rs15092266)\] \[[CrossRef](https://doi.org/10.3390/rs15092266)\]
16. Gnilke, A.; Sanders, T.G.M. Distinguishing Abrupt and Gradual Forest Disturbances with MODIS-Based Phenological Anomaly Series. Front. Plant Sci. **2022**, 13, 863116. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Distinguishing+Abrupt+and+Gradual+Forest+Disturbances+with+MODIS-Based+Phenological+Anomaly+Series&author=Gnilke,+A.&author=Sanders,+T.G.M.&publication_year=2022&journal=Front.+Plant+Sci.&volume=13&pages=863116&doi=10.3389/fpls.2022.863116)\] \[[CrossRef](https://doi.org/10.3389/fpls.2022.863116)\]
17. Guzmán Q., J.A.; Pinto-Ledezma, J.N.; Frantz, D.; Townsend, P.A.; Juzwik, J.; Cavender-Bares, J. Mapping Oak Wilt Disease from Space Using Land Surface Phenology. Remote Sens. Environ. **2023**, 298, 113794. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Mapping+Oak+Wilt+Disease+from+Space+Using+Land+Surface+Phenology&author=Guzm%C3%A1n+Q.,+J.A.&author=Pinto-Ledezma,+J.N.&author=Frantz,+D.&author=Townsend,+P.A.&author=Juzwik,+J.&author=Cavender-Bares,+J.&publication_year=2023&journal=Remote+Sens.+Environ.&volume=298&pages=113794&doi=10.1016/j.rse.2023.113794)\] \[[CrossRef](https://doi.org/10.1016/j.rse.2023.113794)\]
18. Topete-Pozas, C.; Norman, S.P.; Christie, W.M. Multi-Year Hurricane Impacts Across an Urban-to-Industrial Forest Use Gradient. Remote Sens. **2024**, 16, 3890. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Multi-Year+Hurricane+Impacts+Across+an+Urban-to-Industrial+Forest+Use+Gradient&author=Topete-Pozas,+C.&author=Norman,+S.P.&author=Christie,+W.M.&publication_year=2024&journal=Remote+Sens.&volume=16&pages=3890&doi=10.3390/rs16203890)\] \[[CrossRef](https://doi.org/10.3390/rs16203890)\]
19. Wang, F.G.; Xu, Y.J. Hurricane Katrina-Induced Forest Damage in Relation to Ecological Factors at Landscape Scale. Environ. Monit. Assess. **2009**, 156, 491–507. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Hurricane+Katrina-Induced+Forest+Damage+in+Relation+to+Ecological+Factors+at+Landscape+Scale&author=Wang,+F.G.&author=Xu,+Y.J.&publication_year=2009&journal=Environ.+Monit.+Assess.&volume=156&pages=491%E2%80%93507&doi=10.1007/s10661-008-0500-6)\] \[[CrossRef](https://doi.org/10.1007/s10661-008-0500-6)\]
20. Wang, W.T.; Qu, J.J.; Hao, X.J.; Liu, Y.Q.; Stanturf, J.A. Post-Hurricane Forest Damage Assessment Using Satellite Remote Sensing. Agric. For. Meteorol. **2010**, 150, 122–132. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Post-Hurricane+Forest+Damage+Assessment+Using+Satellite+Remote+Sensing&author=Wang,+W.T.&author=Qu,+J.J.&author=Hao,+X.J.&author=Liu,+Y.Q.&author=Stanturf,+J.A.&publication_year=2010&journal=Agric.+For.+Meteorol.&volume=150&pages=122%E2%80%93132&doi=10.1016/j.agrformet.2009.09.009)\] \[[CrossRef](https://doi.org/10.1016/j.agrformet.2009.09.009)\]
21. Feng, Y.L.; Negron-Juarez, R.I.; Chambers, J.Q. Remote Sensing and Statistical Analysis of the Effects of Hurricane Maria on the Forests of Puerto Rico. Remote Sens. Environ. **2020**, 247, 111940. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Remote+Sensing+and+Statistical+Analysis+of+the+Effects+of+Hurricane+Maria+on+the+Forests+of+Puerto+Rico&author=Feng,+Y.L.&author=Negron-Juarez,+R.I.&author=Chambers,+J.Q.&publication_year=2020&journal=Remote+Sens.+Environ.&volume=247&pages=111940&doi=10.1016/j.rse.2020.111940)\] \[[CrossRef](https://doi.org/10.1016/j.rse.2020.111940)\]
22. Dupuis, C.; Lejeune, P.; Michez, A.; Fayolle, A. How Can Remote Sensing Help Monitor Tropical Moist Forest Degradation?—A Systematic Review. Remote Sens. **2020**, 12, 1087. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=How+Can+Remote+Sensing+Help+Monitor+Tropical+Moist+Forest+Degradation?%E2%80%94A+Systematic+Review&author=Dupuis,+C.&author=Lejeune,+P.&author=Michez,+A.&author=Fayolle,+A.&publication_year=2020&journal=Remote+Sens.&volume=12&pages=1087&doi=10.3390/rs12071087)\] \[[CrossRef](https://doi.org/10.3390/rs12071087)\]
23. Helman, D. Land Surface Phenology: What Do We Really “see” from Space? Sci. Total Environ. **2018**, 618, 665–673. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Land+Surface+Phenology:+What+Do+We+Really+%E2%80%9Csee%E2%80%9D+from+Space?&author=Helman,+D.&publication_year=2018&journal=Sci.+Total+Environ.&volume=618&pages=665%E2%80%93673&doi=10.1016/j.scitotenv.2017.07.237)\] \[[CrossRef](https://doi.org/10.1016/j.scitotenv.2017.07.237)\]
24. Huete, A.; Didan, K.; Miura, T.; Rodriguez, E.P.; Gao, X.; Ferreira, L.G. Overview of the Radiometric and Biophysical Performance of the MODIS Vegetation Indices. Remote Sens. Environ. **2002**, 83, 195–213. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Overview+of+the+Radiometric+and+Biophysical+Performance+of+the+MODIS+Vegetation+Indices&author=Huete,+A.&author=Didan,+K.&author=Miura,+T.&author=Rodriguez,+E.P.&author=Gao,+X.&author=Ferreira,+L.G.&publication_year=2002&journal=Remote+Sens.+Environ.&volume=83&pages=195%E2%80%93213&doi=10.1016/S0034-4257\(02\)00096-2)\] \[[CrossRef](https://doi.org/10.1016/S0034-4257\(02\)00096-2)\]
25. Rogan, J.; Schneider, L.; Christman, Z.; Millones, M.; Lawrence, D.; Schmook, B. Hurricane Disturbance Mapping Using MODIS EVI Data in the Southeastern Yucatan, Mexico. Remote Sens. Lett. **2011**, 2, 259–267. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Hurricane+Disturbance+Mapping+Using+MODIS+EVI+Data+in+the+Southeastern+Yucatan,+Mexico&author=Rogan,+J.&author=Schneider,+L.&author=Christman,+Z.&author=Millones,+M.&author=Lawrence,+D.&author=Schmook,+B.&publication_year=2011&journal=Remote+Sens.+Lett.&volume=2&pages=259%E2%80%93267&doi=10.1080/01431161.2010.520344)\] \[[CrossRef](https://doi.org/10.1080/01431161.2010.520344)\]
26. Rossi, E.; Rogan, J.; Schneider, L. Mapping Forest Damage in Northern Nicaragua after Hurricane Felix (2007) Using MODIS Enhanced Vegetation Index Data. Gisci. Remote Sens. **2013**, 50, 385–399. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Mapping+Forest+Damage+in+Northern+Nicaragua+after+Hurricane+Felix+\(2007\)+Using+MODIS+Enhanced+Vegetation+Index+Data&author=Rossi,+E.&author=Rogan,+J.&author=Schneider,+L.&publication_year=2013&journal=Gisci.+Remote+Sens.&volume=50&pages=385%E2%80%93399&doi=10.1080/15481603.2013.820066)\] \[[CrossRef](https://doi.org/10.1080/15481603.2013.820066)\]
27. Wang, F.G.; D’Sa, E.J. Potential of MODIS EVI in Identifying Hurricane Disturbance to Coastal Vegetation in the Northern Gulf of Mexico. Remote Sens. **2010**, 2, 1–18. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Potential+of+MODIS+EVI+in+Identifying+Hurricane+Disturbance+to+Coastal+Vegetation+in+the+Northern+Gulf+of+Mexico&author=Wang,+F.G.&author=D%E2%80%99Sa,+E.J.&publication_year=2010&journal=Remote+Sens.&volume=2&pages=1%E2%80%9318&doi=10.3390/rs2010001)\] \[[CrossRef](https://doi.org/10.3390/rs2010001)\]
28. Gang, C.C.; Pan, S.F.; Tian, H.Q.; Wang, Z.N.; Xu, R.T.; Bian, Z.H.; Pan, N.Q.; Yao, Y.Z.; Shi, H. Satellite Observations of Forest Resilience to Hurricanes along the Northern Gulf of Mexico. For. Ecol. Manag. **2020**, 472, 118243. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Satellite+Observations+of+Forest+Resilience+to+Hurricanes+along+the+Northern+Gulf+of+Mexico&author=Gang,+C.C.&author=Pan,+S.F.&author=Tian,+H.Q.&author=Wang,+Z.N.&author=Xu,+R.T.&author=Bian,+Z.H.&author=Pan,+N.Q.&author=Yao,+Y.Z.&author=Shi,+H.&publication_year=2020&journal=For.+Ecol.+Manag.&volume=472&pages=118243&doi=10.1016/j.foreco.2020.118243)\] \[[CrossRef](https://doi.org/10.1016/j.foreco.2020.118243)\]
29. de Beurs, K.M.; McThompson, N.S.; Owsley, B.C.; Henebry, G.M. Hurricane Damage Detection on Four Major Caribbean Islands. Remote Sens. Environ. **2019**, 229, 1–13. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Hurricane+Damage+Detection+on+Four+Major+Caribbean+Islands&author=de+Beurs,+K.M.&author=McThompson,+N.S.&author=Owsley,+B.C.&author=Henebry,+G.M.&publication_year=2019&journal=Remote+Sens.+Environ.&volume=229&pages=1%E2%80%9313&doi=10.1016/j.rse.2019.04.028)\] \[[CrossRef](https://doi.org/10.1016/j.rse.2019.04.028)\]
30. Turner, H.C.; Galford, G.L.; Lopez, N.H.; Méndez, A.F.; Borroto-Escuela, D.Y.; Ramos, I.H.; González-Díaz, P. Extent, Severity, and Temporal Patterns of Damage to Cuba’s Ecosystems Following Hurricane Irma: MODIS and Sentinel-2 Hurricane Disturbance Vegetation Anomaly (HDVA). Remote Sens. **2023**, 15, 2495. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Extent,+Severity,+and+Temporal+Patterns+of+Damage+to+Cuba%E2%80%99s+Ecosystems+Following+Hurricane+Irma:+MODIS+and+Sentinel-2+Hurricane+Disturbance+Vegetation+Anomaly+\(HDVA\)&author=Turner,+H.C.&author=Galford,+G.L.&author=Lopez,+N.H.&author=M%C3%A9ndez,+A.F.&author=Borroto-Escuela,+D.Y.&author=Ramos,+I.H.&author=Gonz%C3%A1lez-D%C3%ADaz,+P.&publication_year=2023&journal=Remote+Sens.&volume=15&pages=2495&doi=10.3390/rs15102495)\] \[[CrossRef](https://doi.org/10.3390/rs15102495)\]
31. Dronova, I.; Taddeo, S. Remote Sensing of Phenology: Towards the Comprehensive Indicators of Plant Community Dynamics from Species to Regional Scales. J. Ecol. **2022**, 110, 1460–1484. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Remote+Sensing+of+Phenology:+Towards+the+Comprehensive+Indicators+of+Plant+Community+Dynamics+from+Species+to+Regional+Scales&author=Dronova,+I.&author=Taddeo,+S.&publication_year=2022&journal=J.+Ecol.&volume=110&pages=1460%E2%80%931484&doi=10.1111/1365-2745.13897)\] \[[CrossRef](https://doi.org/10.1111/1365-2745.13897)\]
32. Hogan, J.A.; Feagin, R.A.; Starr, G.; Ross, M.; Lin, T.C.; O’connell, C.; Huff, T.P.; Stauffer, B.A.; Robinson, K.L.; Lara, M.C.; et al. A Research Framework to Integrate Cross-Ecosystem Responses to Tropical Cyclones. Bioscience **2020**, 70, 477–489. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=A+Research+Framework+to+Integrate+Cross-Ecosystem+Responses+to+Tropical+Cyclones&author=Hogan,+J.A.&author=Feagin,+R.A.&author=Starr,+G.&author=Ross,+M.&author=Lin,+T.C.&author=O%E2%80%99connell,+C.&author=Huff,+T.P.&author=Stauffer,+B.A.&author=Robinson,+K.L.&author=Lara,+M.C.&author=et+al.&publication_year=2020&journal=Bioscience&volume=70&pages=477%E2%80%93489&doi=10.1093/biosci/biaa034)\] \[[CrossRef](https://doi.org/10.1093/biosci/biaa034)\]
33. Kenneth, R.; Howard, J.; James, P.; Michael, C.; Carl, J. International Best Track Archive for Climate Stewardship (IBTrACS) Project; Version 4; NOAA National Centers for Environmental Information (NCEI): Asheville, NC, USA, 2019. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=International+Best+Track+Archive+for+Climate+Stewardship+\(IBTrACS\)+Project&author=Kenneth,+R.&author=Howard,+J.&author=James,+P.&author=Michael,+C.&author=Carl,+J.&publication_year=2019)\]
34. Zhou, Y.; Matyas, C.J. Regionalization of Precipitation Associated with Tropical Cyclones Using Spatial Metrics and Satellite Precipitation. GISci. Remote Sens. **2021**, 58, 542–561. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Regionalization+of+Precipitation+Associated+with+Tropical+Cyclones+Using+Spatial+Metrics+and+Satellite+Precipitation&author=Zhou,+Y.&author=Matyas,+C.J.&publication_year=2021&journal=GISci.+Remote+Sens.&volume=58&pages=542%E2%80%93561&doi=10.1080/15481603.2021.1908675)\] \[[CrossRef](https://doi.org/10.1080/15481603.2021.1908675)\]
35. Kim, D.; Matyas, C.J. Classification of Tropical Cyclone Rain Patterns Using Convolutional Autoencoder. Sci. Rep. **2024**, 14, 791. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Classification+of+Tropical+Cyclone+Rain+Patterns+Using+Convolutional+Autoencoder&author=Kim,+D.&author=Matyas,+C.J.&publication_year=2024&journal=Sci.+Rep.&volume=14&pages=791&doi=10.1038/s41598-023-50994-5)\] \[[CrossRef](https://doi.org/10.1038/s41598-023-50994-5)\]
36. Abatzoglou, J.T.; Dobrowski, S.Z.; Parks, S.A.; Hegewisch, K.C. TerraClimate, a High-Resolution Global Dataset of Monthly Climate and Climatic Water Balance from 1958–2015. Sci. Data **2018**, 5, 170191. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=TerraClimate,+a+High-Resolution+Global+Dataset+of+Monthly+Climate+and+Climatic+Water+Balance+from+1958%E2%80%932015&author=Abatzoglou,+J.T.&author=Dobrowski,+S.Z.&author=Parks,+S.A.&author=Hegewisch,+K.C.&publication_year=2018&journal=Sci.+Data&volume=5&pages=170191&doi=10.1038/sdata.2017.191&pmid=29313841)\] \[[CrossRef](https://doi.org/10.1038/sdata.2017.191)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/29313841)\]
37. Hargrove, W.W.; Hoffman, F.M. Using Multivariate Clustering to Characterize Ecoregion Borders. Comput. Sci. Eng. **1999**, 1, 18–25. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Using+Multivariate+Clustering+to+Characterize+Ecoregion+Borders&author=Hargrove,+W.W.&author=Hoffman,+F.M.&publication_year=1999&journal=Comput.+Sci.+Eng.&volume=1&pages=18%E2%80%9325&doi=10.1109/5992.774837)\] \[[CrossRef](https://doi.org/10.1109/5992.774837)\]
38. R Core Team. R: A Language and Environment for Statistical Computing; R Core Team: Vienna, Austria, 2024. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=R:+A+Language+and+Environment+for+Statistical+Computing&author=R+Core+Team&publication_year=2024)\]
39. Stekhoven, D.J. missForest: Nonparametric Missing Value Imputation Using Random Forest; Harvard University: Cambridge, MA, USA, 2025. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=missForest:+Nonparametric+Missing+Value+Imputation+Using+Random+Forest&author=Stekhoven,+D.J.&publication_year=2025)\]
40. Kong, D.D.; Zhang, Y.Q.; Gu, X.H.; Wang, D.G. A Robust Method for Reconstructing Global MODIS EVI Time Series on the Google Earth Engine. ISPRS J. Photogramm. Remote Sens. **2019**, 155, 13–24. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=A+Robust+Method+for+Reconstructing+Global+MODIS+EVI+Time+Series+on+the+Google+Earth+Engine&author=Kong,+D.D.&author=Zhang,+Y.Q.&author=Gu,+X.H.&author=Wang,+D.G.&publication_year=2019&journal=ISPRS+J.+Photogramm.+Remote+Sens.&volume=155&pages=13%E2%80%9324&doi=10.1016/j.isprsjprs.2019.06.014)\] \[[CrossRef](https://doi.org/10.1016/j.isprsjprs.2019.06.014)\]
41. Savitzky, A.; Golay, M.J.E. Smoothing + Differentiation of Data by Simplified Least Squares Procedures. Anal. Chem. **1964**, 36, 1627–1639. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Smoothing+++Differentiation+of+Data+by+Simplified+Least+Squares+Procedures&author=Savitzky,+A.&author=Golay,+M.J.E.&publication_year=1964&journal=Anal.+Chem.&volume=36&pages=1627%E2%80%931639&doi=10.1021/ac60214a047)\] \[[CrossRef](https://doi.org/10.1021/ac60214a047)\]
42. Tian, J.Q.; Zhu, X.L.; Chen, J.; Wang, C.; Shen, M.G.; Yang, W.; Tan, X.Y.; Xu, S.; Li, Z.L. Improving the Accuracy of Spring Phenology Detection by Optimally Smoothing Satellite Vegetation Index Time Series Based on Local Cloud Frequency. ISPRS J. Photogramm. Remote Sens. **2021**, 180, 29–44. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Improving+the+Accuracy+of+Spring+Phenology+Detection+by+Optimally+Smoothing+Satellite+Vegetation+Index+Time+Series+Based+on+Local+Cloud+Frequency&author=Tian,+J.Q.&author=Zhu,+X.L.&author=Chen,+J.&author=Wang,+C.&author=Shen,+M.G.&author=Yang,+W.&author=Tan,+X.Y.&author=Xu,+S.&author=Li,+Z.L.&publication_year=2021&journal=ISPRS+J.+Photogramm.+Remote+Sens.&volume=180&pages=29%E2%80%9344&doi=10.1016/j.isprsjprs.2021.08.003)\] \[[CrossRef](https://doi.org/10.1016/j.isprsjprs.2021.08.003)\]
43. Sánchez-Rivera, G.; Gómez-Mendoza, L. Tropical Cyclone Effects on Vegetation Resilience in the Yucatan Peninsula, México, between 2000–2012. Investig. Geográficas **2022**, 203, 203–220. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Tropical+Cyclone+Effects+on+Vegetation+Resilience+in+the+Yucatan+Peninsula,+M%C3%A9xico,+between+2000%E2%80%932012&author=S%C3%A1nchez-Rivera,+G.&author=G%C3%B3mez-Mendoza,+L.&publication_year=2022&journal=Investig.+Geogr%C3%A1ficas&volume=203&pages=203%E2%80%93220&doi=10.14198/INGEO.18499)\] \[[CrossRef](https://doi.org/10.14198/INGEO.18499)\]
44. Basnet, K. Recovery of a Tropical Rain Forest after Hurricane Damage. Vegetatio **1993**, 109, 1–4. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Recovery+of+a+Tropical+Rain+Forest+after+Hurricane+Damage&author=Basnet,+K.&publication_year=1993&journal=Vegetatio&volume=109&pages=1%E2%80%934&doi=10.1007/BF00149540)\] \[[CrossRef](https://doi.org/10.1007/BF00149540)\]
45. Taylor, B.N.; Stedman, E.; Van Bloem, S.J.; Whitmire, S.L.; DeWalt, S.J. Widespread Stem Snapping but Limited Mortality Caused by a Category 5 Hurricane on the Caribbean Island of Dominica. For. Ecol. Manag. **2023**, 532, 120833. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Widespread+Stem+Snapping+but+Limited+Mortality+Caused+by+a+Category+5+Hurricane+on+the+Caribbean+Island+of+Dominica&author=Taylor,+B.N.&author=Stedman,+E.&author=Van+Bloem,+S.J.&author=Whitmire,+S.L.&author=DeWalt,+S.J.&publication_year=2023&journal=For.+Ecol.+Manag.&volume=532&pages=120833&doi=10.1016/j.foreco.2023.120833)\] \[[CrossRef](https://doi.org/10.1016/j.foreco.2023.120833)\]
46. Johnsen, K.H.; Butnor, J.R.; Kush, J.S.; Schmidtling, R.C.; Nelson, C.D. Hurricane Katrina Winds Damaged Longleaf Pine Less than Loblolly Pine. South. J. Appl. For. **2009**, 33, 178–181. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Hurricane+Katrina+Winds+Damaged+Longleaf+Pine+Less+than+Loblolly+Pine&author=Johnsen,+K.H.&author=Butnor,+J.R.&author=Kush,+J.S.&author=Schmidtling,+R.C.&author=Nelson,+C.D.&publication_year=2009&journal=South.+J.+Appl.+For.&volume=33&pages=178%E2%80%93181&doi=10.1093/sjaf/33.4.178)\] \[[CrossRef](https://doi.org/10.1093/sjaf/33.4.178)\]
47. Leitold, V.; Morton, D.C.; Martinuzzi, S.; Paynter, I.; Uriarte, M.; Keller, M.; Ferraz, A.; Cook, B.D.; Corp, L.A.; González, G. Tracking the Rates and Mechanisms of Canopy Damage and Recovery Following Hurricane Maria Using Multitemporal Lidar Data. Ecosystems **2022**, 25, 892–910. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Tracking+the+Rates+and+Mechanisms+of+Canopy+Damage+and+Recovery+Following+Hurricane+Maria+Using+Multitemporal+Lidar+Data&author=Leitold,+V.&author=Morton,+D.C.&author=Martinuzzi,+S.&author=Paynter,+I.&author=Uriarte,+M.&author=Keller,+M.&author=Ferraz,+A.&author=Cook,+B.D.&author=Corp,+L.A.&author=Gonz%C3%A1lez,+G.&publication_year=2022&journal=Ecosystems&volume=25&pages=892%E2%80%93910&doi=10.1007/s10021-021-00688-8)\] \[[CrossRef](https://doi.org/10.1007/s10021-021-00688-8)\]
48. Amaral, C.; Poulter, B.; Lagomasino, D.; Fatoyinbo, T.; Taillie, P.; Lizcano, G.; Canty, S.; Silveira, J.A.H.; Teutli-Hernández, C.; Cifuentes-Jara, M.; et al. Drivers of Mangrove Vulnerability and Resilience to Tropical Cyclones in the North Atlantic Basin. Sci. Total Environ. **2023**, 898, 165413. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Drivers+of+Mangrove+Vulnerability+and+Resilience+to+Tropical+Cyclones+in+the+North+Atlantic+Basin&author=Amaral,+C.&author=Poulter,+B.&author=Lagomasino,+D.&author=Fatoyinbo,+T.&author=Taillie,+P.&author=Lizcano,+G.&author=Canty,+S.&author=Silveira,+J.A.H.&author=Teutli-Hern%C3%A1ndez,+C.&author=Cifuentes-Jara,+M.&author=et+al.&publication_year=2023&journal=Sci.+Total+Environ.&volume=898&pages=165413&doi=10.1016/j.scitotenv.2023.165413)\] \[[CrossRef](https://doi.org/10.1016/j.scitotenv.2023.165413)\]
49. Gong, Y.; Staudhammer, C.L.; Kenney, G.; Wiesner, S.; Zhang, Y.L.; Starr, G. Vegetation Structure Drives Forest Phenological Recovery after Hurricane. Sci. Total Environ. **2021**, 774, 145651. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Vegetation+Structure+Drives+Forest+Phenological+Recovery+after+Hurricane&author=Gong,+Y.&author=Staudhammer,+C.L.&author=Kenney,+G.&author=Wiesner,+S.&author=Zhang,+Y.L.&author=Starr,+G.&publication_year=2021&journal=Sci.+Total+Environ.&volume=774&pages=145651&doi=10.1016/j.scitotenv.2021.145651)\] \[[CrossRef](https://doi.org/10.1016/j.scitotenv.2021.145651)\]
50. Munson, S.M.; Long, A.L. Climate Drives Shifts in Grass Reproductive Phenology across the Western USA. New Phytol. **2017**, 213, 1945–1955. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Climate+Drives+Shifts+in+Grass+Reproductive+Phenology+across+the+Western+USA&author=Munson,+S.M.&author=Long,+A.L.&publication_year=2017&journal=New+Phytol.&volume=213&pages=1945%E2%80%931955&doi=10.1111/nph.14327&pmid=27870060)\] \[[CrossRef](https://doi.org/10.1111/nph.14327)\] \[[PubMed](https://www.ncbi.nlm.nih.gov/pubmed/27870060)\]
51. Norman, S.P.; Koch, F.H.; Hargrove, W.W. Review of Broad-Scale Drought Monitoring of Forests: Toward an Integrated Data Mining Approach. For. Ecol. Manag. **2016**, 380, 346–358. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Review+of+Broad-Scale+Drought+Monitoring+of+Forests:+Toward+an+Integrated+Data+Mining+Approach&author=Norman,+S.P.&author=Koch,+F.H.&author=Hargrove,+W.W.&publication_year=2016&journal=For.+Ecol.+Manag.&volume=380&pages=346%E2%80%93358&doi=10.1016/j.foreco.2016.06.027)\] \[[CrossRef](https://doi.org/10.1016/j.foreco.2016.06.027)\]
52. Liang, X.; Dai, Z.; Mei, X.; Wang, R.; Zeng, W.; Fagherazzi, S. Hurricanes Induced Irreversible Large-Scale Loss of Mangrove Forests. Geophys. Res. Lett. **2025**, 52, e2025GL115692. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=Hurricanes+Induced+Irreversible+Large-Scale+Loss+of+Mangrove+Forests&author=Liang,+X.&author=Dai,+Z.&author=Mei,+X.&author=Wang,+R.&author=Zeng,+W.&author=Fagherazzi,+S.&publication_year=2025&journal=Geophys.+Res.+Lett.&volume=52&pages=e2025GL115692&doi=10.1029/2025GL115692)\] \[[CrossRef](https://doi.org/10.1029/2025GL115692)\]
53. Dahal, D.; Liu, S.; Oeding, J. The Carbon Cycle and Hurricanes in the United States between 1900 and 2011. Sci. Rep. **2014**, 4, 5197. \[[Google Scholar](https://scholar.google.com/scholar_lookup?title=The+Carbon+Cycle+and+Hurricanes+in+the+United+States+between+1900+and+2011&author=Dahal,+D.&author=Liu,+S.&author=Oeding,+J.&publication_year=2014&journal=Sci.+Rep.&volume=4&pages=5197&doi=10.1038/srep05197)\] \[[CrossRef](https://doi.org/10.1038/srep05197)\]
**Figure 1.** Flowchart showing the data sources (in gray) and methodological workflow.
**Figure 1.** Flowchart showing the data sources (in gray) and methodological workflow.
**Figure 2.** The NAB study area showing the 44 hurricane tracks. The rectangles are the focal areas used in the analyses and the colors show the results of clustering.
**Figure 2.** The NAB study area showing the 44 hurricane tracks. The rectangles are the focal areas used in the analyses and the colors show the results of clustering.
**Figure 3.** The fraction of land cover types for the combined hurricane focal areas for each subregion. For detailed land cover by hurricane, see [Table A1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t0A1). Region labels are the same as in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
**Figure 3.** The fraction of land cover types for the combined hurricane focal areas for each subregion. For detailed land cover by hurricane, see [Table A1](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-t0A1). Region labels are the same as in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
**Figure 4.** Regional inter-seasonal and interannual vegetation and climate variation over the full period of record 2001–2022 for the hurricanes included in each subregion. The bold line shows the median value and the boxes show central distribution of the data: (**a**) EVI, (**b**) precipitation, (**c**) temperature, (**d**) vapor pressure deficit.
**Figure 4.** Regional inter-seasonal and interannual vegetation and climate variation over the full period of record 2001–2022 for the hurricanes included in each subregion. The bold line shows the median value and the boxes show central distribution of the data: (**a**) EVI, (**b**) precipitation, (**c**) temperature, (**d**) vapor pressure deficit.
**Figure 5.** The timing of hurricanes with respect to annual land surface phenology (LSP). The curved lines show the 22-year median and first standard deviations of EVI for the hurricane focal areas included in each subregion. Vertical lines show the landfall dates for the hurricanes included in each subregion. The colors are the same as used in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
**Figure 5.** The timing of hurricanes with respect to annual land surface phenology (LSP). The curved lines show the 22-year median and first standard deviations of EVI for the hurricane focal areas included in each subregion. Vertical lines show the landfall dates for the hurricanes included in each subregion. The colors are the same as used in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
**Figure 6.** Percent change in EVI AUC compared to the median wind speed of the hurricane’s focal areas by subregion for (**a**) the first winter after the storm, and (**b**) the first year after the storm. The colors are the same as used in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
**Figure 6.** Percent change in EVI AUC compared to the median wind speed of the hurricane’s focal areas by subregion for (**a**) the first winter after the storm, and (**b**) the first year after the storm. The colors are the same as used in [Figure 2](https://www.mdpi.com/1999-4907/17/4/419#forests-17-00419-f002).
**Figure 7.** Five-year forest (green) and all lands (gray) LSP median EVI profiles for the six case studies centered on the year of the hurricane for (**a**) Charley, (**b**) Laura, (**c**) Harvey, (**d**) Iota, (**e**) Emily, and (**f**) Maria. Landfall dates are shown by red dashed lines.
**Figure 7.** Five-year forest (green) and all lands (gray) LSP median EVI profiles for the six case studies centered on the year of the hurricane for (**a**) Charley, (**b**) Laura, (**c**) Harvey, (**d**) Iota, (**e**) Emily, and (**f**) Maria. Landfall dates are shown by red dashed lines.
**Table 1.** The 44 hurricanes analyzed in this study by subregion. The six storms selected for in-depth analysis as case studies are denoted with an \*. The wind speed corresponds to the median value within the respective track/focal area.
**Table 1.** The 44 hurricanes analyzed in this study by subregion. The six storms selected for in-depth analysis as case studies are denoted with an \*. The wind speed corresponds to the median value within the respective track/focal area.
| | | | | | | | |
|---|---|---|---|---|---|---|---|
| Florida Peninsula (FP) | Central America (CA) | | | | | | |
| Name | Category | Wind speed (km/h) | Landfall | Name | Category | Wind speed (km/h) | Landfall |
| FRANCES | 2 | 157\.42 | 2 September 2004 | BETA | 2 | 133\.34 | 30 October 2005 |
| IRMA | 3 | 178\.71 | 6 September 2017 | IRIS | 3 | 200 | 9 October 2001 |
| JEANNE | 3 | 185\.2 | 16 September 2004 | OTTO | 3 | 180\.57 | 24 November 2016 |
| WILMA | 3 | 185\.2 | 24 October 2005 | ETA | 4 | 175\.94 | 3 November 2020 |
| CHARLEY \* | 4 | 231\.5 | 13 August 2004 | IOTA \* | 4 | 203\.72 | 17 November 2020 |
| IAN | 4 | 207\.42 | 28 September 2022 | FELIX | 5 | 233\.35 | 4 September 2007 |
| Caribbean (C) | Southeast U.S. (SUS) | | | | | | |
| Name | Category | Wind speed (km/h) | Landfall | Name | Category | Wind speed (km/h) | Landfall |
| PALOMA | 2 | 118\.52 | 8 November 2008 | DELTA | 2 | 157\.42 | 9 October 2020 |
| RICHARD | 2 | 148\.16 | 25 October 2010 | GUSTAV | 2 | 171\.31 | 1 September 2008 |
| GRACE | 3 | 185\.2 | 19 August 2021 | IKE | 2 | 162\.97 | 7 September 2008 |
| IAN | 3 | 197\.23 | 27 September 2022 | ISABEL | 2 | 148\.16 | 18 September 2003 |
| SANDY | 3 | 185\.2 | 24 October 2012 | SALLY | 2 | 166\.68 | 16 September 2020 |
| DENNIS | 4 | 213\.9 | 8 July 2005 | DENNIS | 3 | 175\.94 | 10 July 2005 |
| MARIA \* | 4 | 212\.98 | 20 September 2017 | IVAN | 3 | 180\.57 | 16 September 2004 |
| MICHELLE | 4 | 203\.72 | 4 November 2001 | KATRINA | 3 | 194\.46 | 25 August 2005 |
| WILMA | 4 | 185\.2 | 22 October 2005 | RITA | 3 | 175\.01 | 24 September 2005 |
| GUSTAV | 4 | 236\.13 | 30 August 2008 | ZETA | 3 | 185\.2 | 27 October 2020 |
| | | | | IDA | 4 | 212\.98 | 29 August 2021 |
| | | | | LAURA \* | 4 | 219\.46 | 27 August 2020 |
| | | | | MICHAEL | 5 | 259\.28 | 10 October 2018 |
| South Gulf (SG) | West Gulf (WG) | | | | | | |
| Name | Category | Wind speed (km/h) | Landfall | Name | Category | Wind speed (km/h) | Landfall |
| DELTA | 2 | 166\.68 | 7 October 2020 | ALEX | 2 | 166\.68 | 1 July 2010 |
| ERNESTO | 2 | 138\.9 | 8 August 2012 | EMILY | 3 | 185\.2 | 20 July 2005 |
| ISIDORE | 3 | 185\.2 | 22 September 2002 | HARVEY \* | 4 | 194\.46 | 26 August 2017 |
| KARL | 3 | 175\.94 | 17 September 2010 | | | | |
| EMILY \* | 4 | 212\.98 | 18 July 2005 | | | | |
| DEAN | 5 | 235\.2 | 21 August 2007 | | | | |
**Table 2.** Mean percent change in EVI AUC by subregion.
**Table 2.** Mean percent change in EVI AUC by subregion.
| Subregion | 1st Winter | 1st Year |
|---|---|---|
| Florida Peninsula (FP) | −0.02 | −1.25 |
| Central America (CA) | −9.69 | 0\.3 |
| Caribbean (C) | 0\.39 | −1.28 |
| Southeast U.S. (SUS) | −2.58 | −2.78 |
| South Gulf (SG) | 2\.33 | 0\.1 |
| West Gulf (WG) | −0.72 | −9.23 |
**Table 3.** Immediate and secondary response of EVI AUC for forest and all lands for the six hurricane case studies. Units are percent change in the sum of the 16-day EVI values under the winter or annual LSP curve.
**Table 3.** Immediate and secondary response of EVI AUC for forest and all lands for the six hurricane case studies. Units are percent change in the sum of the 16-day EVI values under the winter or annual LSP curve.
| Cases | All Lands | Forest | | |
|---|---|---|---|---|
| 1st Winter | 1st Year | 1st Winter | 1st Year | |
| Charley (FP) | −2.32 | 2\.74 | −2.47 | 1\.17 |
| Laura (SUS) | −7.88 | −9.43 | −10.23 | −13.64 |
| Harvey (WG) | −9.47 | −9.83 | −2.25 | −4.72 |
| Iota (CA) | −23.53 | −5.17 | −35.84 | −2.96 |
| Emily (SG) | 6\.44 | −6.63 | 11\.4 | −4.9 |
| Maria (C) | −7.85 | 0\.9 | −1.12 | 1\.64 |
| | |
|---|---|
| | **Disclaimer/Publisher’s Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the [Creative Commons Attribution (CC BY) license](https://creativecommons.org/licenses/by/4.0/).
## Share and Cite
**MDPI and ACS Style**
Topete-Pozas, C.; Norman, S.P. Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022). *Forests* **2026**, *17*, 419. https://doi.org/10.3390/f17040419
**AMA Style**
Topete-Pozas C, Norman SP. Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022). *Forests*. 2026; 17(4):419. https://doi.org/10.3390/f17040419
**Chicago/Turabian Style**
Topete-Pozas, Carlos, and Steven P. Norman. 2026. "Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022)" *Forests* 17, no. 4: 419. https://doi.org/10.3390/f17040419
**APA Style**
Topete-Pozas, C., & Norman, S. P. (2026). Land Surface Phenology Reveals Region-Specific Hurricane Impacts Across the North Atlantic Basin (2001–2022). *Forests*, *17*(4), 419. https://doi.org/10.3390/f17040419
Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details [here](https://www.mdpi.com/about/announcements/784).
## Article Metrics |
| Shard | 31 (laksa) |
| Root Hash | 11974975279136771031 |
| Unparsed URL | com,mdpi!www,/1999-4907/17/4/419 s443 |