Investing The Impacts Of Frozen Soil Representation On High-Resolution Streamflow Simulation At The Continental Scale
Presenter: Jetal Agnihotri1
Co-Author(s): Ahmad Tavakoly, Matthew Geheran
Advisor(s): Dr. Guo-Yue Niu and Dr. Ali Behrangi
1Department of Hydrology and Atmospheric Sciences, University of Arizona
Panapto Presentation Video
Oral Session 2: Data-Driven and Physically-Based Modeling
Frozen ground affects the timing and magnitude of streamflow in the melting season as it impedes infiltration of snowmelt and rainfall. Different representations of frozen soil schemes and its effects on river discharge simulations in snowmelt dominated gauges are investigated in this study. We employ the Noah-MP land surface model to evaluate simulations using supercooled soil water (two options) and soil hydraulic properties (conductivity and matrix potential: 4 options). With the gridded surface and subsurface runoff resulting from Noah-MP as inputs, we use the Routing Application for Parallel computatIon of Discharge (RAPID) river routing model to produce daily discharge at 52 USGS gauges in the Mississippi River Basin (MRB) for the period of 2015 to 2019. In addition, Noah-MP is driven by two gridded precipitation products using the North American Land Data Assimilation System (NLDAS-2) precipitation and the Integrated Multi-satellitE Retrievals for GPM (IMERG) final products. Mean Kling-Gupta Efficiency (KGE) values show that improvement in the range of 20% - 57% is attained when models with hydraulic conductivity scheme that produces higher infiltration are used to simulate discharge at gauges in MRB. These effects are particularly evident over regions dominated by frozen soil. In contrast, models with different supercooled soil water scheme did not show significant improvements as the soil hydraulic schemes. Overall, models forced with IMERG-Final precipitation produces more accurate discharge simulations compared to models forced by NLDAS precipitation at most gauges.