Operational water forecast ability of the HRRR-iSnobal combination: An evaluation to adapt into production environments

Authors: MEYER, JOACHIM - University Of Utah; HOREL, JOHN - University Of Utah; KORMOS, PATRICK - National Weather Service; Hedrick, Andrew; TRUJILLO, ERNESTO - Boise State University; SKILES, MCKENZIE - University Of Utah

Submitted to: Geoscientific Model Development
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/12/2022
Publication Date: 1/10/2023
Citation: Meyer, J., Horel, J., Kormos, P., Hedrick, A., Trujillo, E., Skiles, M. 2023. Operational water forecast ability of the HRRR-iSnobal combination: An evaluation to adapt into production environments. Geoscientific Model Development. 16(1):233-250. https://doi.org/10.5194/gmd-16-233-2023.
DOI: https://doi.org/10.5194/gmd-16-233-2023

Interpretive Summary: Freshwater resupply originating from seasonal snow in mountain regions is changing in amount and aerial extent. To estimate the amount of water, water-resources forecasters use data from the past, which are increasingly unrepresentative and causing errors with the current prediction methods. This work presented an alternative approach for water-resource forecasting from seasonal snow and evaluated its performance. The alternative method was tested in a watershed in the Western United States and compared against measurements taken within the region, surveys from an airplane, and how it fared against the current established prediction method. The results achieved were promising, and the alternative method captured the amounts of snow until the peak season was reached. However, once the snow started to melt, the alternative method became less accurate than the measurements. This discrepancy can be addressed by revisiting estimations internal to the new method. Even with some improvements recommended, this new method has a high potential to improve the current predictions of water forecasters. Adaptation into current water forecaster environments would improve resilience to the ongoing and upcoming changes to seasonal snow and help ensure the accuracy of freshwater resupply forecasts going forward.

Technical Abstract: Operational water-resource forecasters, such as the Colorado Basin River Forecast Center (CBRFC) in the Western United States rely on historical records to calibrate the temperature-index models currently used for snowmelt runoff predictions. This data dependence is increasingly challenged, with global and regional climatological factors changing the seasonal snowpack in mountain watersheds. To evaluate and improve the CBRFC modeling options, this work ran the physically based snow energy balance iSnobal model, forced with outputs from the High-Resolution Rapid Refresh (HRRR) numerical weather model across four years in a subset region. Compared to in-situ, remotely sensed, and the current operational CBRFC model, the iSnobal-HRRR coupling showed well-reconstructed snow depths until peak accumulation (Mean differences between -0.20 and +0.28 m). Once snowmelt set in, iSnobal-HRRR showed that simulated snowmelt was slower relative to observations, depleting snow on average up to 34 days later. The melting period is a critical component for water forecasting. Based on the results, there is a need for revised energy balance calculations in iSnobal, which is a recommended future improvement to the model. Nevertheless, the presented performance and architecture make iSnobal-HRRR a promising combination for the CBRFC production needs, where there is a demonstrated change to the seasonal snow in the mountain ranges around the Colorado River Basin. Long term goal is to introduce the iSnobal-HRRR coupling in day-to-day CBRFC operations, and this work created the foundation to expand and evaluate larger domains.