Accounting for field-scale heterogeneity in the ecohydrological modeling of large arid river basins: Strategies and relevance

It remains a great challenge to represent field-scale (e.g., meters to tens of meters) heterogeneity in the ecohydrological modeling of large river basins due to the tradeoff between spatial resolution and computational cost. This study improved an existing ecohydrological model, HEIFLOW, by introducing the subgrid structure of land cover, multilayer soil water simulation and accurate spatial coverage of irrigation within the grid. These improvements enable the model to provide reliable simulations over a wide spectrum of spatial scales, from the field scale to the large-basin scale (i.e., 10 to 10 km). The new model was implemented in the Heihe River Basin, the second largest endorheic river basin in China, with a grid size of 1 km by 1 km for a modeling domain of approximately 90,589 km. The major study findings include the following. First, in arid areas with sparse vegetation, ignoring the subgrid characteristics of land surfaces will lead to significant errors that may be further propagated when the modeling results are used to support management or scaled up for larger-scale climate modeling. Second, the multilayer soil structure can improve ecohydrological simulations in terms of temporal variations, and it is necessary to separate a thin surface layer from the soil zone in arid areas. In the case study, a single-layer soil structure would introduce greater than 10% error in simulating the annual maximum leaf area index (LAI). Third, considering the accurate spatial coverage of irrigation within the grid cell is critical for successful simulations of ecohydrological processes in arid areas. In the case study, accurate spatial coverage would lead to −31%, +46%, and +13% changes in the simulated average soil evaporation, transpiration and LAI, respectively, over the entire area with irrigation. Overall, this study provides a unique perspective to address the scale issue in ecohydrological modeling and reveals the importance of field-scale heterogeneity to the management of water resources and ecosystems based on ecohydrological modeling.