FLAME: a novel approach for modelling burned area in the Brazilian biomes using the Maximum Entropy concept - Input Data

This repository contains driving data used by training and evaluation of FLAME in the "FLAME: a novel approach for modelling burned area in the Brazilian biomes using the Maximum Entropy concept" paper. All NetCDF files are on regular, 0.5-degree grids on a monthly timestep over Brazil. Not all variables were used in the final analysis NetCDF File Variable Used/not Used Source/Reference burned_area.nc Burned area As training data MCD64A1/ Giglio et al. (2018) burned_area_nat_veg.nc Burned area in natural vegetation As training data MCD64A1/ Giglio et al. (2018) and Mapbiomas, 2022 burned_area_non_nat_veg.nc Burned area in non natural vegetation As training data MCD64A1/ Giglio et al. (2018) and Mapbiomas, 2022 tas_max.nc Maximum Temperature Used ISIMIP3a FRIELER et al. (2023) precip.nc Precipitation Used vpd.nc Vapor pressure deficit Not Used rhumid.nc Relative Humidity Not Used consec_dry_days.nc Consecutive number of dry days Not Used soilM.nc Soil Moisture Not Used JULES-ES lightn.nc Lightning Not Used ISIMIP3a FRIELER et al. (2023) popDen.nc Population density Not Used road_density.nc Road density Used GRIP global (MEIJER et al., 2018) cveg.nc Vegetation carbon Not Used JULES-ES csoil.nc Carbon in dead vegetation Used JULES-ES forest.nc Forest Used MAPBIOMAS, 2022 grassland.nc Grassland Not Used savanna.nc Savanna Not Used cropland.nc Cropland Not Used pasture.nc Pasture Used np.nc Number of patches Not Used Calculated from MAPBIOMAS,2022 ed.nc Edge density Used