Equal-area spline functions applied to a legacy soil database to create weighted-means maps of soil organic carbon at a continental scale

There is a growing need for raster-based soil data to support modelling at regional and continental scales. The GlobalSoilMap consortium aims to satisfy this need with the production of a suite of digital soil maps of various soil properties at six standard depths for most of the land surface of the Earth. Initially, the maps will be produced using legacy soil data (soil data already available). In the United States, the State Soil Geographic (STATSGO2) database is a rich source of legacy soil information, but the STATSGO2 map is vector-based and soil components' soil property data is organised by horizon. We therefore applied the equal-area spline function to the soil components of STATSGO2 map units in order to obtain estimates of soil organic carbon (SOC) content at the GlobalSoilMap standard depth increments. Using these estimates, we calculated the weighted mean of SOC for each STATSGO2 map unit at each GlobalSoilMap.net depth increment and gridded these weighted means at 100 m resolution for the contiguous United States. The result is a set of maps of SOC content for each GlobalSoilMap depth increment accompanied by an indication of the within-map unit variability. In addition, we show how the use of other “metadata maps” is essential for avoiding misunderstandings about reported values and discuss some of the limitations of the approach.

► We apply equal-area splines to soils of the U.S. STATSGO2 database.
► We use splines to estimate soil organic carbon at GlobalSoilMap depth increments.
► We map soil organic carbon content for the contiguous U.S. from spline estimates.
► We demonstrate how use of metadata maps augments comprehension of reported values.