Spatial datasets of wildfire risk to rangeland carbon for the conterminous United States (30m) circa 2014 and 2020 landscapes

Wildfire risk to rangeland carbon provides estimates of the total standing biomass and carbon in rangelands of the conterminous United States (CONUS) circa 2014 and 2020, as well as calculations of carbon emissions from wildfire and carbon remaining following wildfire. Rangelands are non-forested ecosystems including shrublands and grasslands, covering approximately 821 million acres of CONUS as of 2014, and 833 million acres as of 2020. We modeled standing biomass of rangelands through multiple steps. Shrub biomass was modeled from LANDFIRE (https://landfire.gov) data layers representing shrub cover, shrub height, and shrub type. Herbaceous biomass, including grasses and forbs, was calculated from Rangeland Analysis Platform (https://rangelands.app/) data. Litter and duff were modeled using data from the Fuel Characteristic Classification System Fuelbeds dataset. Total rangeland biomass was then used in the First Order Fire Effects Model (FOFEM) to calculate standing carbon and biomass, carbon and biomass remaining following wildfire, and carbon emissions conditional upon wildfire occurring. By utilizing burn probability outputs from the Large Fire Simulator (FSim), carbon emissions conditional upon wildfire were converted to annual expected carbon emissions given burn probability (wildfire likelihood). Data are provided as individual georeferenced TIFF files, with values reported in tons per acre.Geospatial quantification of wildfire risk to rangeland carbon can provide public and private land managers with information about how best to allocate limited budgets to mitigate risks to a valuable resource. Carbon market managers may use this information to inform their required carbon buffer pool contributions.