A physically-based, distributed-parameter hydrologic model was used to simulate a recent flood event in the city of Hafr Al Batin, Saudi Arabia to gain a better understanding of the runoff generation and spatial distribution of flooding. The city is located in a very arid catchment. Flooding of the city is influenced by the presence of three major tributaries that join the main channel in and around the heavily urbanized area. The Integrated Multi-satellite Retrievals for Global Precipitation Measurement Mission (IMERG) rainfall product was used due to lack of detailed ground observations. To overcome the heavy computational demand, the catchment was divided into three sub-catchments with a variable model grid resolution. The model was run on three subcatchments separately, without losing hydrologic connectivity among the sub-catchments. Uncalibrated and calibrated satellite products were used producing different estimates of the predicted runoff. The runoff simulations demonstrated that 85% of the flooding was generated in the urbanized portion of the catchments for the simulated flood. Additional model simulations were performed to understand the roles of the unique channel network in the city flooding. The simulations provided insights into the best options for flood mitigation efforts. The variable model grid size approach allowed using physically-based, distributed models—such as the Gridded Surface Subsurface Hydrologic Analysis (GSSHA) model used in this study—on large basins that include urban centers that need to be modeled at very high resolutions.

Continuous monitoring of stream water quality is needed as it has significant impacts on human and ecological health and well-being. Estimating water quality between sampling dates requires model simulation based on the available geospatial and water quality data for a given watershed. Models such as the Soil and Water Assessment Tool (SWAT) can be used to estimate the missing water quality data. In this study, SWAT was used to estimate water quality at a monitoring station near the outlet of the San Antonio River. Precipitation data from both rain gauges and weather radar were used to force the SWAT simulations. Virtual rain gauges which were based on weather radar data were created in the approximate centres of the 163 sub-watersheds of the San Antonio River Basin for SWAT simulations. This method was first tested in a smaller watershed in the middle of the Guadalupe River Basin resulting in increased model efficiency in simulating surface run-off. The method was then applied to the San Antonio River watershed and yielded good simulations for surface run-off (R² = 0.7), nitrate (R² = 0.6) and phosphate (R² = 0.5) at the watershed outlet (Goliad, TX – USGS (United States Geological Survey) gauge) as compared to observed data. The study showed that the proper use of weather radar precipitation in SWAT model simulations improves the estimation of missing water quality data.