H111-0001
Evaluation of National Water Model in simulating Evapotranspiration fluxes across multiple spatio-temporal scales

Friday, 11 December 2020
Poster
Amir Hossein Mazrooei1, Arezoo Rafieeinasab1, Aubrey L Dugger1, Prasanth Valayamkunnath2, David Gochis3, Xia Feng4 and Brian A. Cosgrove5, (1)National Center for Atmospheric Research, Boulder, CO, United States, (2)National Center for Atmospheric Research, Boulder, VA, United States, (3)NCAR, Boulder, CO, United States, (4)NOAA/NWS/NWC, Tuscaloosa, AL, United States, (5)NOAA/NWS/OHD, Silver Spring, MD, United States
Abstract:
Enhancing the performance of the operational National Water Model (NWM) in simulating Evapotranspiration (ET) fluxes can help to further reduce the biases in streamflow predictions. In this study we evaluate the NWM’s skill in simulating ET fluxes across CONUS against Ameriflux in situ point-observations as well as remotely sensed observations of ET from MODIS satellite. Daily ET simulations from two versions of NWM (i.e., v.2.0 and v.2.1) are utilized to explore various verification metrics such as bias, %bias, NSE and KGE during different seasons and for low/medium/high flow events. The validation against 85 Ameriflux towers is performed using the NWM grids (at 1km resolution) that encloses tower locations. The validation against MODIS ET is performed based on the spatially averaged ET over more that 1000 basins located in various hydroclimatic conditions. The NWM hourly outputs are post-processed to daily (8-day) integrated estimates to compare with Ameriflux (MODIS) observations. Furthermore, the NWM’s sensitivity to ET-related parameters such as root depth under vegetated landscapes (e.g., grasslands, pasture, irrigated croplands) is analysed over the high plains region of the continental U.S. where the NWM has shown a persistent positive bias in the baseflow component of total streamflow predictions.