Multi-Model Predictions of Design Low Flow Conditions in the Great Salt Lake Basin Using NextGen Water Resources Modeling Framework

Abstract Submission: Accurate estimation of low flow conditions may be extremely challenging in ungauged basins in regions such as Great Salt Lake (GSL). We can estimate the design low flow conditions based on 7Q10 flow using the USGS streamflow observations in the gauged basins. However, streamflow predictions in the ungauged basins remain challenging as there is limited hydrologic modeling capability. With the advent of the Next Generation Water Resources Modeling Framework (alias NextGen), we are able to integrate multiple hydrologic models into the NextGen framework for predicting streamflow at over 2.7 million river reaches, with 99.7% of the reaches being ungauged. Among many capabilities, the NextGen framework simplifies the implementation and evaluation of various model structures, including (i) Conceptual Functional Equivalent (CFE) to the National Water Model (NWM), (ii) TOPMODEL, and (iii) Catchment Arid/Semi-Arid Model (CASAM), which allows us to investigate the accuracy of multiple model structures with different hydrological process representations (e.g., surface/subsurface runoff, arid/semi-arid climate) in predicting the low flow conditions in the watersheds (e.g., GSL). In the NextGen framework, we calibrate and validate multiple hydrologic models for the GSL basin using the Dynamically Dimensioned Search (DDS) calibration scheme. Our investigation shows that CFE 2.0 outperforms CFE 1.0 whereas CASAM would outperform all in predicting streamflow. However, this study paves the way to investigate how these models perform in predicting the design low flow (i.e., 7Q10) that is typically used as a standard practice so we can provide appropriate design low flow estimates for the water supply planning and management as well as for ensuring water quality in the GSL basin at the gauged and ungauged locations.

Learning Objectives/Expected Outcome (Optional) :