Coupling the CSM-CROPGRO-Soybean crop model with the ECOSMOS Ecosystem Model – An evaluation with data from an AmeriFlux site

• Published in: Agricultural and forest meteorology 2023-11, Vol.342, p.109697, Article 109697
• Main Authors: Dias, Henrique Boriolo, Cuadra, Santiago Vianna, Boote, Kenneth J., Lamparelli, Rubens Augusto Camargo, Figueiredo, Gleyce Kelly Dantas Araújo, Suyker, Andrew E., Magalhães, Paulo Sérgio Graziano, Hoogenboom, Gerrit
• Format: Article
• Language: English
• Subjects: agroecosystems; carbon; carbon dioxide; climate; CO2 exchange; Crop modelling; crop models; decision support systems; ecological models; Eddy covariance; energy balance; Evapotranspiration; forests; genotype; Glycine max (L.) Merrill; Growth; irrigation; meteorology; model validation; Nebraska; phenology; soil water balance; soybeans; surface water
• ISSN: 0168-1923
• DOI: 10.1016/j.agrformet.2023.109697

•The CROPGRO-Soybean model was coupled with the ECOSMOS ecosystem model.•The model was evaluated with flux, soil water, and crop observations.•Good simulations of mass, energy and water exchanges were obtained.•Soybean phenology and growth were most accurately simulated by the model.•The model accurately simulated soybean yield at harvest. Process-based simulation models, such as land surface (LSM) and crop models, are useful tools for studying the impacts of the environment, management and genotype on agricultural production. LSM are capable of simulating crop development and growth, but not in as much detail as the processes embedded in crop models. Crop models, on the other hand, do not usually have the ability to solve the surface water, energy and carbon balances, which can help to assess the feedbacks between climate and agricultural systems. The goals of this study were first to implement the well-known Cropping System Model (CSM)-CROPGRO-Soybean model of the Decision Support System for Agrotechnology Transfer (DSSAT) into the Ecosystem Model Simulator (ECOSMOS) LSM and then to evaluate this coupling to simulate surface fluxes and crop performance of irrigated and rainfed soybean agroecosystems with comprehensive data from an AmeriFlux site. After model coupling, simulations were benchmarked against multiple flux (carbon dioxide, energy, and water), phenology, growth, and yield observations obtained from field scale experiments across 14 seasons from irrigated and rainfed fields near Mead, Nebraska. Calibration of the physiological parameters of ECOSMOS and of the CROPGRO-Soybean genetic coefficients was conducted. Common metrics were employed to assess model performance. Overall, the coupled model reproduced both the magnitude and seasonal patterns of the energy balance, carbon dioxide exchange, evapotranspiration and soil water balance. The crop model coupling into ECOSMOS preserves the known high skill of the CROPGRO-Soybean model in simulating soybean phenology and growth dynamics. The simulated yields were consistent with the observations at field level. Although the ECOSMOS-CROPGRO-Soybean model is sufficiently robust to simulate surface fluxes and crop performance of soybean agroecosystems at field scale for the environments of eastern Nebraska, further evaluation for a wide range of climatic and soil conditions and management practices is warranted.