Carbon and Water Budgets in Multiple Wheat-Based Cropping Systems in the Inland Pacific Northwest US: Comparison of CropSyst Simulations with Eddy Covariance Measurements

Accurate carbon and water flux simulations for croplands are greatly dependent on high quality representation of management practices and meteorological conditions, which are key drivers of the surface-atmosphere exchange processes. Fourteen site-years of carbon and water fluxes were simulated using the CropSyst model over four agricultural sites in the inland Pacific Northwest (iPNW) US from October 1, 2011 to September 30, 2015. Model performance for field-scale net ecosystem exchange of CO2 (NEE) and evapotranspiration (ET) was evaluated by comparing simulations with long-term eddy covariance measurements. The model captured the temporal variations of NEE and ET reasonably well with an overall r of 0.78 and 0.80, and a low RMSE of 1.82 g C m(-2) d(-1) and 0.84 mm d(-1) for NEE and ET, respectively. The model slightly underestimated NEE and ET by 0.51 g C m(-2) d(-1) and 0.09 mm d(-1), respectively. ET simulations showed better agreement with eddy covariance measurements than NEE. The model performed much better for the sites with detailed initial conditions (e.g., SOC content) and management practice information (e.g., tillage type). The CropSyst results showed that the winter wheat fields could be annual net carbon sinks or close to neutral with the net ecosystem carbon balance (NECB) ranging from 92 to -17 g C m(-2), while the spring crop fields were net carbon sources or neutral with an annual NECB of -327 to -3 g C m(-2). Simulations for the paired tillage sites showed that the no-till site resulted in lower CO2 emissions for the crop rotations of winter wheat-spring garbanzo, but had higher carbon loss into the atmosphere for spring canola compared to the conventional tillage site. Water budgets did not differ significantly between the two tillage systems. Winter wheat in the high-rainfall area had higher crop yields and water use efficiency but emitted larger amounts of CO2 into the atmosphere than in the low-rainfall area. Based on model evaluations in this study, CropSyst appears promising as a tool to simulate field-scale carbon and water budgets and assess the effects of different management practices and local meteorological conditions for the wheat-based cropping systems in this region.