You are here:
Problems and Prospects of SWAT Model Application on an Arid/Semi-arid Watershed in Arizona
Citation:
Yuan, Y., W. Nie, AND E. Sanders. Problems and Prospects of SWAT Model Application on an Arid/Semi-arid Watershed in Arizona . Presented at SEDHYD 2014 Joint Conference, Reno, NV, March 23 - 27, 2014.
Impact/Purpose:
Given the growing demand for water due to urban growth and the likelihood of decreasing precipitation due to climate change, water sustainability has become a dominant issue in the arid and semi-arid regions such as the Southwestern USA. To address a nation’s or a region’s water-related sustainability problems, one of key elements is to characterize and quantify water resources for different future scenarios including different Landuse and Landcover (LULC) and climate in order to develop a better management practice. In the arid and semiarid areas, it has been a great challenge to quantify the water resources due to limited access and monitoring systems on the land and limited capability of hydrological and water quality models to handle the unique hydrology associated with these regions. The Soil and Water Assessment Tool (SWAT) was designed to evaluate the impact of LULC change on watershed hydrology and water quality and has been widely applied for watershed scenario analysis, its application in the arid and semi-arid regions has been few. Thus, the overall objective of this study is to evaluate SWAT applicability on an arid and semi-arid watershed in Arizona to support sustainable water resources development.
Description:
Hydrological characteristics in the semi-arid southwest create unique challenges to watershed modelers. Streamflow in these regions is largely dependent on seasonal, short term, and high intensity rainfall events. The objectives of this study are: 1) to analyze the unique hydrology of a watershed located in the southwestern USA; and 2) to evaluate the hydrologic model, Soil and Water Assessment Tool (SWAT)’s applicability on this watershed. USGS historical precipitation and stream discharge patterns were analyzed to determine the hydrological characteristics of the upper San Pedro watershed. It was found that runoff was decreased in downstream gauging locations because of transmission loss due to low groundwater level. Based on this analysis, SWAT model was calibrated to reflect the unique hydrological characteristics of the watershed. After calibration, the Nash-Sutcliffe efficiency (NSE) and the coefficient of determination (R2) values were above 0.5 for monthly and annual runoff simulation for 3 gauges (except NSE for annual calibration at Redington gage), and the percent bias (PBIAS) values were within the range of ±25% (except annual calibration at Charleston gage), suggesting satisfactory model performance. Through the calibration process, several parameters including Baseflow aphla factor (days), SCS curve number for moisture condition II, Soil evaporation compensation factor, threshold depth of water in the shallow aquifer for ‘revap’ to occur and available soil water capacity that are sensitive to discharge in this watershed were identified and they could be suggested for reflecting the hydrological characteristics of other arid/semi-arid watersheds.