Soil Functional Mapping: Using Data-Model Integration to Improve Regional-Scale SOC Forecasts

With advances in the understanding of the mechanisms leading to the persistence or vulnerability of soil organic carbon (SOC) at the profile scale, it is essential to develop infrastructure to integrate this knowledge with landscape-scale mapping and models. To address this need, we are developing a soil functional unit framework, intended to better scale mechanistic soil knowledge by merging geospatial datasets with targeted sample collection and analyses. Here we provide a proof of concept of this approach for SOC stocks (the soil function of interest) in the East River study area located near Gothic, Colorado, USA. We first generate a map estimating SOC stocks based only on available geospatial datasets, including factors such as topography, vegetation, geology, and basic soil maps. We then compare the mapped functional units against an independent SOC dataset of 450 soil profiles (~1700 samples) collected from the study region and refine the soil functional map to best capture the spatial variability observed in the dataset.  With the calibrated soil functional unit mapping algorithm, we can then calculate SOC stocks at landscape scales and better constrain the mechanisms that drive the observed heterogeneity.  The resulting data-driven soil functional maps can then be merged with regional scale SOC models to enhance forecasts of SOC change in response to disturbances.