Global wildfire exposure dataset

Based on the FireTracks Scientific Dataset, the global wildfire exposure dataset implements space-time alignment between fire clusters and climate as well as socioeconomic data across the globe from 2002 to 2020. The global wildfire exposure dataset is produced to meet the demand for wildfire exposure and its analysis of global inequity. The data structure is shown in the following: fire_clusters.csv: Contains detailed records of global wildfire clusters, including fire characteristics, geographic distribution, and population exposure. Column Description Units Valid Range Data Type cp fire component label - >=0 int64 duration fire duration days >= 1 uint16 maxFRP_sum sum of maximum fire radiative powers MW*10 >= 0 float64 area total burned area km^2 >= 0.86 (1 MODIS pixel) float64 country country of occurrence - - string continent continent of occurrence - - string dtime_min ignition date (YYYY-MM-DD) - >= 2002-01-01 datetime64 lat_mean mean location latitude degrees [-180, 180] float64 lon_mean mean location longitude degrees [-90, 90] float64 population primary population exposure to wildfire persontime >=0 int64 5km_pop secondary population exposure to wildfire persontime >=0 int64 year year of occurrence - [2002, 2020] int64 month month of occurrence - [1, 12] int64 exposure_four_covariate.csv: Aggregates population exposure and covariates for equity-focused analyses, linking wildfire impacts to socioeconomic and climatic drivers. Column Description Units Valid Range Data Type population population exposed to wildfire persontime >=0 int64 year year of occurrence - [2002, 2020] int64 country country of occurrence - - string continent continent of occurrence - - string gdp National Gross Domestic Product for the corresponding year million U.S. dollars >=0 float64 country_pop national population of the corresponding year persontime >=0 int64 fwi Fire Weather Index - >=0 float32 vpd Vapor Pressure Deficit kPa [0, 10] float32 ndvi Normalized Difference Vegetation Index - [0, 1] float64 dnbr Delta Normalized Burn Ratio - [-2, 2] float64 iso3 standardized country code for cross-dataset integration - - string cluster results from K-means clustering - [0, 2] int64