RothC is an established model for predicting soil organic carbon (SOC) changes in response to environmental conditions and management. The model lacks algorithms for carbon input estimation and a differentiated consideration of carbon input sequestration from organic amendments commonly used in agriculture. Moreover, it does not consider the higher stability of below-ground crop residues in relation to above-ground crop residues.

RothC was combined with two empirical approaches for quantifying carbon inputs from above- and below-ground crop residues and tested on 439 SOC data series from 36 arable long-term field experiments in Central and Northern Europe. Effects of carbon input quality on model fit were quantified with linear mixed models based on the analytical solution of RothC. Model parameters that describe the stability of incoming carbon were calibrated using a multi-site approach and Bayesian calibration. A second calibration study combined the determination of partitioning of incoming carbon into pools with different turnover rates and model responses to temperature and soil water content. With this calibration we showed that the contribution of above-ground residues to SOC is lower than when estimated with default RothC paramerization. We also show that the relative contribution from roots to SOC is higher than that from above-ground residues. Moreover, the degradability of organic amendments was highly variable between amendment categories and increased for all model configurations in the following order: farmyard manure < farmyard compost < sawdust < sewage sludge < peat. The proposed RothC partition coefficients for above-ground residues, roots and several commonly used organic amendments as estimated for this large data set should be useful for other studies in temperate climates. We also show that the analytical solution of RothC is closely related to the much simpler empirical humus balancing approaches used by farm advisory services. This provides opportunities to build bridges between the more process-oriented SOC models used in research and the well established instruments used by the farming community to assess the effects of agricultural management practices on SOC changes.