Evaluation and Application of the Community Land Model for Simulating Energy and Carbon Exchange in Agricultural Ecosystems

University of Minnesota Ph.D. dissertation. June 2016. Major: Land and Atmospheric Science. Advisor: Timothy Griffis. 1 computer file (PDF); vii, 131 pages.Land surface models (LSMs) serve as important tools for studying the interactions between the atmosphere and ecosystems, understanding biophysical feedback processes, and predicting future climate. Most state-of-the-art LSMs assessed in the fifth phase of the Coupled Model Inter-comparison project (CMIP5), however, did not include process-based crop models with comprehensive physiology and phenology. One of the most widely used LSMs, the Community Land Model (CLM), has only recently included prognostic crops (CLM4-Crop). This thesis research evaluated the CLM4-Crop LSM for corn and soybean agro-ecosystems using 54 site-years of data, improved its representation of phenology for application within the US Corn Belt, and assessed CLM’s performance in simulating carbon and energy exchange for the conterminous United States. The results indicated that CLM4-Crop has a biased phenology during the early growing season and underestimated carbon emissions from corn and soybean sites. The carbon budget for the conterminous United States estimated by CLM4-Crop showed a greater carbon sink (-0.191±0.020 PgC yr-1) compared to the original CLM4-CN model (-0.139±0.034 PgC yr-1). A modified model, CLM4-CropM, with improved phenology and crop physiological parameters was developed to better simulate crop gross primary production (GPP), ecosystem respiration (ER) and leaf area index (LAI). The key parameters include the maximum carboxylation rate at 25 oC (Vcmax25), specific leaf area at the top of the canopy (slatop) and a parameter determining the initial carbon allocation to leaves (fleafi). CLM4-CropM estimated a smaller carbon sink (-0.094±0.023 PgC yr-1) than either CLM4-CN or CLM4-Crop, and the magnitude and trend of NEE was in good agreement with remote sensing-based estimates. This research has important implications for studying the long-term carbon budget of intensively cultivated regions