LAUSR

Abstract Soil particle density (ρS) is one of the basic physical properties of soils and represents an essential element of diverse pedotransfer functions (PTFs). Recent findings support a PTF where the solid soil particles were considered as a mass-based mixing ratio of two components, soil mineral substance (SMS) and soil organic matter (SOM), which are related to the corresponding component particle densities. In this study, we developed a hierarchically structured model considering the subcomponents of SMS and SOM that are based on soil characteristics recorded by routine soil inventories - soil texture and soil organic carbon (SOC). Regarding SOM, we were faced with the problem that both the SOC to SOM conversion factor and the SOM particle density are variable. Finally, we generated a mechanistic approach to predict the particle density of soils as affected by subfractions of both soil mineral substances and soil organic matter. For the model calibration, related literature provided the required boundary conditions. We calibrated our model on a dataset (N = 501) from locations worldwide covering the full range of possible soil organic matter contents, diverse textures and soil parent materials. In the results, we quantified i) the SOC dependency of the SOC to SOM conversion factor as well as of the SOM subfraction’s particle densities and ii) the mean particle densities of the clay-, silt- and sand-size fractions.