Submitted to: American Society of Agricultural Engineers Meetings Papers
Publication Type: Proceedings
Publication Acceptance Date: 4/1/1995
Publication Date: N/A
Citation: N/A Interpretive Summary: Considerable resources are needed to experimentally study the impacts of agricultural management practices on production and the environment. This limitation necessitates the use of mathematical (computer) models to enhance our understanding of the complex natural environment. In this study we examined the ability of the USDA Root Zone Water Quality Model (RZWQM) to simulate dryland/no-till corn production systems in eastern Colorado by comparing model simulated results with field measurements. The intentions were to gauge model usefulness and to evaluate the effects of crop surface residue on evaporation and soil water conservation. In this report, model limitations and strenghts are identified and recommednations for model improvement are discussed. Grain production was over-predicted on summit and sideslopes and under-predicted on toeslope positions due to errors in predicting runoff (runon) water contribution from the upslope positions. In nRZWQM, detailed rainfall duration and intensity data are needed to accurately predict runoff and runon. This becomes particularly important if using models to assess the economic consequences (i.e., graing yield) of dryland cropping systems since every unit of stored water translates to additional yield and thus profit. In regard to water conservation, the model suggest that in dryland no-till farming under semi-arid eastern Colorado with infrequent rainfall, surface residue control of soil evaporation is minimal. In light of this, the remaining question is "what are the dominant residue-related mechanisms that contribute to increased soil water storage."
Technical Abstract: Field tests of models of agricultural production systems are needed to gauge their usefulness. In this study, we evaluated the ability of the Root Zone Water Quality Model to simulate dryland/no-till cropping systems in eastern Colorado. We focused on the corn phase of the wheat-corn-fallow system where model results were compared with measured seasonal soil water, ,crop grain yield/biomass production and N uptake. The model was first parameterized using data from one site. Independent evaluation runs, using the parameterized model were conducted comparing model results with measurements from different soils at two sites. Based on this, we conclude that the overall performance of the model is satisfactory although some enhancements may improve model results. Because of rainfall intensity/ duration data given as hourly totals rather than actual, the rainfall/ runoff process in the model predicted small runoff. This led to under- predicting soil water and thus production on the lowland toeslope position which receive significant amounts of surface runoff from the upslope positions. Actual rainfall intensity data is needed to adequately predict runoff and production. The model suggests that under dryland no-till farming in semi-arid eastern Colorado with infrequent rainfall, retardation of soil evaporation by flat surface residue is minimal, mainly because of the process being more soil- than energy-limiting. The model was found to have educational utility and provided a tool to synthesize knowledge and identify knowledge gaps.