Ensemble model and input model rasters for soil organic carbon stock mean and uncertainties for Alaska, British Columbia, Washington, Oregon, California, and Hawai'i

This data publication contains products of the USDA Forest Service, Pacific Northwest Research Station (PNW) Carbon Science Synthesis project as part of the Carbon Dynamics Initiative. The geographic area of interest for the work includes all ecoregions in Alaska, British Columbia, Washington, Oregon, California, and Hawai'i. The raster files (GeoTiffs) included in this publication represent baseline estimates of soil organic carbon (SOC) derived from available databases for two standard depth ranges (0-30 and 0-100 centimeters) and represent the time period 2008-2021. This publication includes both the input and output files for this project. Input files include the collected SOC model estimates, from 11 different sources, that were used to develop ensemble model estimates. Output files include the SOC stock average estimates (megagrams of carbon per hectare [Mg C/ha]), estimates of uncertainty in the form of standard deviation of input model estimates (Mg C/ha), and estimates of data intensity (number of estimates per pixel). Mean SOC, deviation of model inputs, and intensity values are from input models that were not correlated to each other within a given ecoregion. Correlations were systematically tested for each model compared to all other models in a pairwise process. For correlations between model pairs, the overlapping estimates from input models with the coarsest spatial scale, or smallest coverage area were removed.
Data were collected to develop baseline estimates for SOC and associated uncertainties across the ecoregions of Alaska, British Columbia, Washington, Oregon, California and Hawai'i to aid in the goal of the Carbon Science Synthesis project to determine the magnitude and distribution of carbon pools across our area of interest.