RapidFEM4D: aboveground biomass density maps for post-Hurricane Ian forest monitoring in Florida

dataset

posted on 2025-04-04, 15:22 authored by Inacio BuenoInacio Bueno, Carlos Alberto Silva, Caio Hamamura, Monique Bohora Schlickmann, Victoria M. Donovan, Jeff AtkinsJeff Atkins, Kody Brock, Jinyi Xia, Denis ValleDenis Valle, Jiangxiao QiuJiangxiao Qiu, Jason Vogel, Andres Susaeta, Ajay Sharma, Mauro Karasinski, Carine Klauberg

This dataset provides spatially aboveground biomass density (AGBD) maps from 2022 to 2024, created to assess the impacts of Hurricane Ian on forest ecosystems in Florida. The Hurricane Ian made landfall in Florida on September 28, 2022, as a powerful Category 4 and was the second major storm of the 2022 Atlantic hurricane season.

Field data were collected in situ during the spring of 2023 and 2024 across 27 plots, each measuring 25 × 25 meters. These plots were strategically distributed to represent a full gradient of AGBD. Within each plot, tree species were identified, and diameter at breast height (DBH) and height measurements were recorded for all individuals with DBH greater than 10 cm. AGBD was estimated for each tree based on species-specific allometric equations, incorporating species identity, DBH, and tree height.

To generate the AGBD maps, we integrated multiple remote sensing datasets, including GEDI L4A AGBD estimates, Harmonized Landsat-Sentinel (HLS) optical imagery, and Sentinel-1C radar data, along with other ancillary layers, totaling 245 predictor variables. A random forest regression model was trained using field data and predictor layers to estimate AGBD, which was then upscaled to create a continuous AGBD map for the study area.

The final AGBD prediction for each year was obtained by averaging results from multiple bootstrap iterations.. Additionally, uncertainty maps were derived by calculating the standard deviation of predictions across these iterations, providing insight into spatial variability and model confidence.

Resources in this dataset:

AGBD predictions for consecutive years: RapidFEM4D_AGBD_prediction_2022.tif; RapidFEM4D_AGBD_prediction_2023.tif, and RapidFEM4D_AGBD_prediction_2024.tif
Map uncertainties for the respective AGBD predictions: RapidFEM4D_AGBD_uncertainty_2022.tif; RapidFEM4D_AGBD_uncertainty_2023.tif, and RapidFEM4D_AGBD_uncertainty_2024.tif
Field data used for technical validation: RapidFEM4D_field_data.xlsx

Funding

RapidFEM4D: A web-based mapping platform for rapidly assessing the impacts and near-term recovery of Hurricane Ian on forest ecosystems in Florida

National Institute of Food and Agriculture

Find out more...

History

Data contact name

Bueno, Inacio

Data contact email

ithomazbueno@ufl.edu

Publisher

Ag Data Commons

Intended use

The AGBD maps and field data serve as calibration, validation, and direct comparison datasets for studies on post-hurricane forest recovery, carbon assessment, disaster response, and ecosystem resilience. Additionally, the dataset supports land management decisions, tracks long-term forest dynamics, assesses climate change impacts, and improves carbon sequestration models.

Use limitations

Uncertainty maps highlight areas where predictions may be less reliable due to model limitations, particularly in regions with sparse field data or complex forest structures. Users should interpret these areas with caution when applying the data to localized analyses. The 30-meter resolution of these maps imposes constraints on the level of detail that can be captured. Fine-scale variations in biomass distribution may be smoothed or underrepresented, limiting the dataset’s applicability for studies requiring high-resolution biomass estimation. Additionally, small forest gaps, individual trees, and fine-scale structural differences may not be fully detected, which could lead to discrepancies when comparing with higher-resolution datasets, such as UAV or airborne LiDAR-based biomass estimates.

Temporal Extent Start Date

2022-03-01

Temporal Extent End Date

2024-05-31

Frequency

annually

Theme

Geospatial

Geographic Coverage

{"type":"FeatureCollection","features":[{"type":"Feature","geometry":{"type":"Polygon","coordinates":[[[-82.852432,25.183],[-80.031362,25.183],[-80.031362,30.42415],[-82.852432,30.42415],[-82.852432,25.183]]]},"properties":{}}]}

Geographic location - description

Florida, United States

ISO Topic Category

disaster
economy
environment

National Agricultural Library Thesaurus terms

aboveground biomass; forests; monitoring; Florida; data collection; hurricanes; forest ecosystems; spring; trees; allometry; equations; tree height; remote sensing; radar; algorithms; regression analysis; prediction; uncertainty; standard deviation; models; Internet; carbon; ecological resilience; land management; forest dynamics; climate change; carbon sequestration; canopy gaps; unmanned aerial vehicles