Evaluation and inversion of a net ecosystem carbon exchange model for grasslands and croplands

A one-dimensional soil water, heat, and CO2 flux model (SOILCO2), a pool concept of soil carbon turnover (RothC), and a crop growth module (SUCROS) was coupled to predict the net ecosystem exchange (NEE) of carbon. This model, further referred to as AgroC, was extended with routines for managed grassland as well as for root exudation and root decay. In a first step, the coupled model was applied to two winter wheat sites and one upland grassland site in Germany. The model was calibrated based on soil water content, soil temperature, biometric, and soil respiration measurements for each site, and validated in terms of hourly NEE measured with the eddy covariance technique. The overall model performance of AgroC was acceptable with a model efficiency >0.78 for NEE. In a second step, AgroC was optimized with the eddy covariance NEE measurements to examine the effect of various objective functions, constraints, and data-transformations on estimated NEE, which showed a distinct sensitivity to the choice of objective function and the inclusion of soil respiration data in the optimization process. Both, day and nighttime fluxes, were found to be sensitive to the selected optimization strategy. Additional consideration of soil respiration measurements improved the simulation of small positive fluxes remarkably. Even though the model performance of the selected optimization strategies did not diverge substantially, the resulting annual NEE differed substantially. We conclude that data-transformation, definition of objective functions, and data sources have to be considered cautiously when using a terrestrial ecosystem model to determine carbon balances by means of eddy covariance measurements.