Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Etats Unis → 2020 → Application of a Semi-Distributed Hydrologic Model to Understand Hydrologic Flux Partitioning of an Arid, Highly Managed Watershed

University of California Riverside (2020)

Application of a Semi-Distributed Hydrologic Model to Understand Hydrologic Flux Partitioning of an Arid, Highly Managed Watershed

Jha, Aarushi

Titre : Application of a Semi-Distributed Hydrologic Model to Understand Hydrologic Flux Partitioning of an Arid, Highly Managed Watershed

Auteur : Jha, Aarushi

Université de soutenance : University of California Riverside

Grade : Master of Science (MS) in Environmental Science 2020

Irrigated agriculture accounts for 70% of global freshwater withdrawals and increases in the frequency and intensity of droughts and population growth have exerted significant stress on freshwater supply in recent decades. There is a growing need to improve the quantification of hydrologic fluxes in agricultural ecosystems and reduce water use particularly in arid and semi-arid regions where irrigation contributes substantially to the region’s water budget. Given the limited availability of hydrometric observations at catchment scale, hydrologic models have been widely used in research and operational applications. Recent advances in computational resources and availability of spatially distributed information have increased the use of numerical models for water balance estimation at large scale. The objective of this study is to quantify hydrologic flux partitioning of the Salton Sea Watershed, a 21,801 km2 highly managed watershed in the arid regions of Southern California and Mexico. We apply a semi-distributed hydrologic model, Soil and Water Assessment Tool, to simulate hydrologic processes of two distinct hydrologic systems - undisturbed headwater subbasins in the mountain ranges and irrigated agricultural subbasins in the southern part of the watershed. The model setup incorporates a large endorheic lake (Salton Sea) and major irrigation canals to distribute imported Colorado River inflows across major irrigation regions. To reduce computational demand, we demonstrate a method for selecting optimal spatial discretization of Hydrologic Response Units while capturing spatial variability of hydrologic fluxes. Model calibration for 8 undisturbed headwater subbasins is conducted on a monthly time step using the SUFI-2 optimization algorithm. We found that the suggested calibration methodology for maintaining spatial variability of parameters did not yield satisfactory model performance using Nash-Sutcliff Efficiency as an objective function (NSE < 0.5). An alternative calibration approach of replacing values of spatially distributed parameters improved model performance for 4 out of the 8 headwater subbasins (NSE > 0.5) during calibration and validation periods where streamflow was perennial and mean annual streamflow was larger than 60mm/year. Model calibration for 13 agricultural subbasins was conducted using a combination of manual and automatic calibration techniques for monthly streamflow at two major rivers draining to the Salton Sea. The calibrated model did not yield satisfactory results for either rivers (NSE < 0.5). However, temporal dynamics of streamflow is well captured by the model (linear correlation coefficient R > 80%). We performed two simulation scenarios to assess the impacts of changes in the Colorado River inflows on the water balance. We found that under these scenarios, changes in the Salton Sea inflows were directly caused by seasonal fluctuations in the Colorado River inflows, and evapotranspiration and irrigation were not significantly impacted compared to the baseline simulation even for the low flow scenario. These findings have important implications for water and agricultural management in the Salton Sea Watershed where competing water demand exist among agriculture and urban users, and natural habitats. Future efforts will focus on improving the calibration approach by incorporating the remotely sensed evapotranspiration and soil moisture products and representing subsurface hydrologic processes using physically based modeling approaches.

Présentation et version intégrale

Page publiée le 2 décembre 2021