2026-10-12T05:10:49Z https://www.scidb.cn/oai

10.57760/sciencedb.38866 2026-06-08T11:35:43Z

2026-06-08 Spatial patterns and temporal dynamics of world carbon emissions from 1700-2100 doi:10.57760/sciencedb.38866 en This dataset comprises a synthesized compilation of global wildfire carbon dioxide emissions (WCEs) data, integrating historical reconstructions (1700–2000), contemporary satellite observations (2001–2020), and future projections (2021–2100), alongside country-level WCE trends for the top ten forest-rich nations . Historical data are derived from an ensemble of Fire Model Intercomparison Project (FireMIP) Dynamic Global Vegetation Models (DGVMs), harmonized using standardized CRU TS climate forcing and historical land-use change inputs, with validation against paleoenvironmental proxies (charcoal, ice-core black carbon). Contemporary data aggregate four satellite-derived products (GFEDv4s, GFASv1.2, FEERv1.0, QFEDv2.6) aligned to a 0.1°×0.1° grid, processed via ensemble averaging to reduce single-product bias. Future projections combine outputs from fire-enabled DGVMs (e.g., CLM4.5, ORCHIDEE-SPITFIRE), CMIP5/6 Earth System Models, and machine learning ensembles, covering Shared Socioeconomic Pathways (SSPs: SSP1-2.6, SSP2-4.5, SSP5-8.5). Temporally, the dataset spans 1700–2100 with annual resolution (sub-annual for contemporary satellite data); spatially, it includes global gridded data (0.1°–0.25° resolution) and country-level statistics for Russia, Canada, the USA, Brazil, China, Democratic Republic of Congo, Australia, Indonesia, and others. Trend.xlsx consist of 301-year historical, 20-year contemporary, and 80-year future WCE time series (units: Pg CO₂ yr⁻¹), with 12–15 records per model/scenario. Fraction.xlsx include 20-year national WCE trends (units: Pg CO₂ yr⁻¹) for 7 countries. Missing values arise from historical model uncertainties (e.g., limited representation of human-fire interactions pre-1850), contemporary satellite gaps in detecting small agricultural fires or cloud-obscured events, and future scenario divergences. Inter-dataset discrepancies of up to 40% stem from variations in satellite detection sensitivity, combustion completeness assumptions (e.g., 0–80 m burning depth in peatlands vs. grasslands), emission factors, and fire-climate-human interaction parameterizations. The dataset is provided in NetCDF (gridded) and CSV (national/timeseries) formats, compatible with standard GIS (QGIS, ArcGIS) and statistical software (R, Python), with full metadata available upon request from the corresponding authors. Wildfire carbon dioxide emissions; Spatial patterns; temporal dynamics Yu Liang Tianxiao Ma Bo Liu RESTRICTED dataset Science Data Bank