Connecting (near) saturated hydraulic conductivity to soil aggregation and carbon content on croplands across Europe

The soil hydraulic conductivity function describes the soil’s ability to transmit water and in land surface models (LSMs) this function is often estimated from soil texture and soil organic carbon (SOC) content. In addition, soil hydraulic conductivity in the wet range (K_sat and K@h=-2cm) is strongly influenced by soil structure, whereby structural information is currently not used to estimate the soil’s hydraulic conductivity. Neglect of soil structure can therefore lead to false estimations of soil hydrological fluxes and soil water storage. In order to quantify the impact of soil structure on the soil’s hydraulic conductivity, we conducted infiltration measurements and analysed aggregate size distribution as an indicator of soil structure in croplands along a climate and organic carbon gradient across Europe. We found that soil aggregation was controlled by a combination of the log transformed SOC and clay content, as well as by the log transformed soil moisture. When information on the mean weighted diameter (MWD) of the sand-free aggregates and the mass distribution between aggregate size fractions was included in linear models that predicted K_sat or K@h=-2cm, the predictions had a 13.3 % lower RMSE and could explain up to 67 % of variation in the K_sat data. When the information was not included, the models could only explain 56 % of the variation in the K_sat data, in general predictions were better for K_sat than K@h=-2cm. For K@h=-2cm, the mass distribution between sand-free aggregate size was more important than the MWD of the aggregates, which was not the case for K_sat. For K_sat, the most important predictors were the interaction of the MWD of the aggregates with the carbon and clay content, the interaction of the carbon and clay content and the soil moisture and soil temperature at the time of field measurement. The significant contribution of indicators of soil structure in these models confirm that soils with larger aggregate MWD, indicating a more developed soil structure, have higher K_sat.  Therefore, soil aggregation should be taken into account, either by including the MWD of soil aggregates as a variable or developing proxies, in order to estimate water partitioning in soils to ultimately incorporate into LSMs.