Location: Hydrology and Remote Sensing LaboratoryTitle: Combined use of crop yield statistics and remotely sensed products for enhanced simulations of evapotranspiration within an agricultural watershed
|LEE, S. - University Of Maryland|
|QI, JUNYU - University Of Maryland|
|YANG, YANG - US Department Of Agriculture (USDA)|
|KWAK, D. - National Institute Of Forest Science|
|KIM, H. - University Of Virginia|
|LAKSHMI, V. - University Of Virginia|
Submitted to: Agricultural Water Management
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/20/2022
Publication Date: 1/22/2022
Citation: Lee, S., Qi, J., McCarty, G.W., Anderson, M.C., Yang, Y., Zhang, X., Moglen, G.E., Kwak, D., Kim, H., Lakshmi, V. 2022. Combined use of crop yield statistics and remotely sensed products for enhanced simulations of evapotranspiration within an agricultural watershed. Agricultural Water Management. 264:107503. https://doi.org/10.1016/j.agwat.2022.107503.
Interpretive Summary: r, assess, and predict conservation practices because of their capability to simulate cumulative long-term impacts of conservation practices across multiple spatial scales. One major concern regarding any hydrologic modeling exercise is predictive uncertainty. Although the reliability of simulated outcomes is confirmed via model calibration and validation to some degree, predictive uncertainty always exists. The primary goal of this study is to explore the SWAT model predictive uncertainty in estimating ET for a cropland-dominant watershed within the coastal plain of the Chesapeake Bay Watershed. Hydrologic and biophysical parameters were tuned against streamflow, evapotranspiration (ET), and annual crop yield. Various sets of model parameter inputs were tested and only one parameter set produced acceptable performance for streamflow, ET, and annual crop yield. Most parameter sets failed to satisfactorily depict crop yield although they accurately simulated ET. This finding indicated that model uncertainty (so-called “equifinality”) in estimates of ET was substantially reduced with the inclusion of crop yield as a biophysical constraint. Robust model calibration requires biophysical parameters. Inclusion of irrigation practices should also lead to stronger estimates of ET. This study demonstrated that the model predictive uncertainty can be reduced by including these additional constraints.
Technical Abstract: Remote sensing-driven evapotranspiration products (RS-ET) have been widely adopted as an additional constraint on watershed modeling to enhance the model prediction while reducing predictive uncertainty. However, biophysical parameters for estimating evapotranspiration (ET) in watershed models have been poorly calibrated without the use of biophysical constraints. The goal of this study is to assess the Soil and Water Assessment Tool (SWAT) predictive uncertainty in estimates of ET depending on the inclusion or exclusion of annual crop yield used as biophysical constraints. We compared simulated results with acceptable performance measures for two calibration data types (streamflow and RS-ET) and those with acceptable performance measures for three calibration data types (streamflow, RS-ET, and crop yield). Kling-Gupta Efficiency (KGE) was used as the performance measure for both streamflow and RS-ET, and crop yield performance was quantified using Percent-bias (P-bias). 17 of 3,000 parameter sets met pre-defined performance measures for streamflow and RS-ET. high uncertainty because of natural and anthropogenic factors (e.g., cultivation practices). Remotely sensed evapotranspiration products (RS-ET) have been adopted as an additional constraint on watershed modeling to enhance the accuracy of water cycling predictions while reducing predictive uncertainty. However, plant parameters affecting evapotranspiration (ET) in watershed models are poorly calibrated without the use of appropriate constraints. The goal of this study is to assess the predictive uncertainty of the Soil and Water Assessment Tool (SWAT), depending on the inclusion or exclusion of annual crop yield as an additional constraint for an agricultural watershed. We analyzed the simulated results with acceptable performance measures depending on a varying degree of model constraints: one constraint (streamflow), two constraints (streamflow and RS-ET) and three constraints (streamflow, RS-ET, and crop yield). As the model constraint increased, the number of acceptable parameter sets were substantially reduced from 180 (acceptable for streamflow) to 116 (acceptable for streamflow and RS-ET) and 2 (acceptable for streamflow, RS-ET, and crop yield). In addition, overall model performance measures for ET were greatest in the simulation results with three constraints. The parameter set with the best ET performance measures was also acceptable for predicting crop yield. Based on these results, we conclude that that crop yield data can be adopted as a model constraint for agricultural watersheds to reduce of model uncertainty in ET simulations and to increase of model prediction accuracy.