Evaluating Alternative Baseflow Estimation Methods for Improving National Water Model Forecasting

Session: CUAHSI Summer Institute at the National Water Center
Chair: Jerad Bales (CUAHSI)
The 2018 CUAHSI Summer Institute (SI) was held June 10 – July 26, 2018 at the National Water Center on the Tuscaloosa campus of the University of Alabama. Twenty-five graduate students, two course coordinators, nine faculty theme leaders, and three CUAHSI staff participated, as well as a number of guests from the NWS, USGS, and elsewhere. Students at the 2018 SI focused on hyper-resolution modeling, groundwater – surface connections, data from volunteer monitoring, and river hydraulics. Students used National Water Model outputs and observations to explore potential improvements to the NWM and to describe uncertainty. Among the data sets utilized was the CUAHSI-curated Hurricane Harvey hydrometeorological data from 2017. The session will begin with a presentation from Dr. Thomas Graziano, Chief of the NWS Office of Water Prediction, followed by presentations by selected 2018 students.

\"Evaluating Alternative Baseflow Estimation Methods for Improving National Water Model Forecasting\"
Speaker: Joseph M. Krienert (Southern Illinois University)
Co-Authors: Minki Hong (Texas A&M University), Ritesh Karki (Auburn University), and Sama S. Memari (University of Alabama)

The National Water Model (NWM) became fully operational in August of 2016. This hydrologic model uses a framework of real time climatic data and observed physiographic attributes to produce forecasts of surficial fluvial systems across the Contiguous US. Comparison of the NWM streamflow estimations with USGS observed datasets has shown that the NWM estimates are not entirely accurate, especially during low flow conditions. A potential influence on this miscalculation lies in the NWM’s conceptual (not physicallyexplicit) estimation of baseflow to streams, and this non-linear conceptual baseflow model only expresses a part of the interaction between groundwater and surficial hydrology.

This research evaluates the current representation groundwater discharge in the NWM with a case study of five watersheds located within the Northern High Plains region. A comparison between USGS observed stream flow, a systematically calibrated groundwater model (Peterson et al 2016), and the output hindcasts of the NWM will be used to evaluate the current model’s fitness in representing surface water - groundwater interaction in gaining streams; with a particular focus on baseflow estimation. Based on the results of this analysis, formulation of alternative functions representing the relationship between subsurface storage and groundwater discharge will be tested for potential improvements in future updates of the NWM.