2020 CSDMS meeting-004
Quantifying Land-Use Impacts on SOC, Soil Texture, and Bulk Density and simulating Global Human-Induced Soil Degradation from 850 to 2015
Pei-Ling Wang, Geography, University of Victoria Victoria , Canada. firstname.lastname@example.org
Johannes Feddema, Geography, University of Victoria Victoria , Canada. email@example.com
The characteristics of soils control the influence of how land use and land cover (LULC) change the global water, energy, and biogeochemical cycles. Plant health, and the exchange of energy, water and biogeochemical components at the surface interface is partly controlled by soil properties. Different soil types modify vegetation responses to existing climate forcings, and each soil type also responds differently to the same land-use practice. Currently, Earth System Models often use single soil columns with averaged properties and the same properties stay constant over time regardless of LULC changes. This leads to uncertainties in assessing LULC impacts. To improve the estimates of land surface change in Earth System Models, we build a soil degradation model to compute annual soil properties from 850 to 2015. The model includes three parts: first, to quantify human LULC impacts, we collected 1099 observations from 174 published literature of human impacts of agriculture, pasture, grazing, and vegetation harvest on soil organic carbon (SOC), texture, and bulk density. Under each LULC unit, we defined the combined impact of LULC, management, climate (represented by NPP or moisture index), and soil texture on each soil property based on observations and regression models. In the second part, we link an existing LULC dataset to four hydrologic soil groups from 850 to 2015, based on demonstrated soil preferences for eight LULCs under current conditions. We conclude that humans prefer hydrologic soil groups (HSGs) in order from B, D, C, to A (generally from high to low silt content). This ranking was applied to construct the history of LULC on each soil type at the half-degree grid resolution. Results primarily distribute croplands to HSG B in 850, while HSG A has the most undisturbed area. Over time, preferred soils (HSGs B and D) experience increased use for cropland areas, while poor soils (HSGs C and A) are occupied predominantly by increasing areas in grazing land and secondary non-forests. Finally, based on the established LULC and soil relations from 850 to 2015, we altered soil properties in each soil group according to global variations of environmental factors to model human-induced soil degradation. Vertical and temporal variations are applied based on observations. Results demonstrate how soil degradation occurs under historical LULCs and provide better land surface characteristics to improve Earth systems modeling.