Journal of Advances in Modeling Earth Systems Глобальные и планетарные изменения

Downscaling HRRR Radiation for Distributed Snow Modeling: Part 2—Topographic Correction of Incoming Longwave and Evaluation of Combined Radiation Updates

Alvaro Robledano, Matthew Olson, Joachim Meyer, J. Michelle Hu, S. McKenzie Skiles
Wiley · 2026-11-01 · DOI 10.1029/2025ms005609

Аннотация

Abstract The spatially distributed, process‐based snow model iSnobal is a longstanding research model that is widely used in the Western United States. The model has undergone development to be forced with outputs from the High‐Resolution Rapid Refresh (HRRR) numerical weather prediction model. However, the longwave radiation forcing remains internally prepared, with incoming longwave parameterized using empirical relationships that do not fully represent interactions in mountainous terrain. To unify model inputs, in this second paper of a two‐part series, incoming longwave radiation from HRRR is downscaled and adjusted for topography to inform the incoming longwave forcing in iSnobal. Evaluation in the East River Watershed, Colorado, during water years 2022–2023 shows that, relative to the traditional treatment of incoming longwave, this update reduces RMSE on average by 36.6–44.5 Wm −2 at measurement sites. Basin‐wide, when combined with the shortwave radiation updates from Part 1, the net radiative balance increased by +3 Wm −2 , due to reduction in net solar radiation (−34 Wm −2 ) and increase in net longwave radiation (+37 Wm −2 ). These changes lengthened the snow melt‐out date bias at in situ observation sites by 5–9 days, preserving the spatial variability of modeled snow depth across the watershed. The combined updates align radiation forcings with other model forcings, improve the ability of iSnobal to model processes contributing to radiative heterogeneity in complex terrain, and maintain accurate predictions of snowmelt in mountain watersheds for both research and operational applications.

Читать у издателя →