MicroBasin - A High-resolution Watershed Scale Ecohydrologic Model for Agro-ecosystems

Monday, 15 December 2014
Fidel Maureira1, Claudio O Stockle1, Mingliang Liu1, Roger Nelson1, Erin S Brooks2 and Rolf Sommer3, (1)Washington State University, Pullman, WA, United States, (2)University of Idaho, Biological and Agricultural Engineering, Moscow, ID, United States, (3)Consultative Group on International Agricultural Research (CGIAR), Montpellier, France
For comprehensive agro-ecosystem management, a model to track vertical and lateral flows of water, nutrients and carbon over the landscape and small watersheds is a highly attractive tool to evaluate ex ante, changes that result from the pressure of adaptation and mitigation to climate change. For this purpose, we developed a high-resolution watershed-scale ecohydrologic model (MicroBasin) that simulates the interactions between crops, soil and atmosphere systems over the vertical and horizontal space domain. This sub-model for crop growth and soil biogeochemical processes is derived from the well-developed CropSyst model, which simulates crop yields and soil organic matter processes under given environmental conditions (water, nitrogen, and heat, etc.) and management (tillage, fertilization, irrigation, rotation, and harvest). For vertical movement of water and nutrients along the soil profile, both hourly cascade and finite difference infiltration methods were implemented. For horizontal flows, MicroBasin simulates the overland flows and within-soil lateral flows according to the roughness of land surface, watershed topology, and soil properties. We considered a closed energy budget incorporating frozen soil and snowpack processes with detailed characterization of specific parameters, with the freeze-thaw process simulated in each soil layer. Exhaustive observations from two small watersheds of the Inland Pacific Northwest of the United States were used for evaluating this model. Over the watershed with non-tillage (1.1 ha), the estimated evapotranspiration had a root mean square error (RMSE) of 0.94 mm, whereas the simulated net exchange ecosystem had a RMSE of 2.31 gC/m2 over two observation years. Over the watershed with conventional tillage (1.9 ha), the modeled hourly and daily water fluxes (such as runoff) and soil moisture showed a consistency with observations (RMSE= 54 m3/yr, R2=0.46).