Table 1 Open data sources used for dataset preparation.
From: Wildfire spreading prediction using multimodal data and deep neural network approach
Data source | Acquisition period | Spatial resolution | Units | Format | Description |
|---|---|---|---|---|---|
MODIS MCD12Q1 v06120 | 2021 year | 500 m | Classes | GeoTIFF | Land cover: annual International Geosphere–Biosphere Programme classification |
Copernicus GLO-30 digital elevation model21 | 2011–2015 years | 30 m | m | GeoTIFF | Elevation |
WorldPop population density22 | 2020 year | 30 arc-seconds | Number of people per km2 | GeoTIFF | Population density |
MODIS MOD15A2H v06123 | 8 days before fire initiation | 500 m | % | GeoTIFF | Fraction of photosynthetically active radiation |
m2/m2 | Leaf area index | ||||
MODIS MOD13A1 v06124 | 16 days before fire initiation | 500 m | – | GeoTIFF | Normalized difference and enhanced vegetation index |
MODIS MOD16A2 v00625 | 8 days before fire initiation | 500 m | kg/m2 | GeoTIFF | Evapotranspiration and potential evapotranspiration |
MODIS MOD11A1 v06126 | For every prediction day | 1 km | K | GeoTIFF | Daytime and nighttime land surface temperature |
ERA5-Land27 | For every prediction day | 9 km | m/s | NetCDF | Eastward and northward components of the 10 m wind |
K | Temperature in the atmosphere | ||||
m | Accumulated liquid and frozen water, comprising rain and snow, that falls to the Earth’s surface |
