Authors: Anna Grunes, Kate Hale, Joachim Meyer, Beverley Wemple – University of Vermont
Title: Improved Snowmelt Modeling in the Northeastern United States
Presentation Type: Poster Presentation
Abstract: Montane, seasonally snow-covered regions of the Northeastern United States are understudied, sensitive to warming winter conditions, and prone to mid-winter rain-on-snow (ROS) driven runoff and flooding. Historical trends in the regional snowpack and recent hazardous events, including a significant December 2023 ROS event, highlight the need to revisit operational snow modeling that improves snowmelt-derived runoff contributions across event-to-annual temporal scales in this region. We evaluate the performance of a physically based snow mass and energy balance model, iSnobal, over the Winooski River Basin in north-central Vermont, a 2800 km2 domain spanning elevations from the Green Mountains (1340 m) to Lake Champlain (30 m). We force iSnobal with High-Resolution Rapid Refresh (HRRR) numerical weather model data to assess its suitability to supplement operational forecasting.
Across water years 2023-2025, representing a range of winter climate conditions, iSnobal well reproduced seasonal snowpack evolution and melt timing in headwater regions of the Winooski, with variable biases in snow water equivalent. At both the seasonal and event scale, we addressed negative temperature biases in HRRR, which impacts precipitation phase partitioning and therefore overestimates seasonal snowpack accumulation, by conducting systematic sensitivity tests. Our findings exemplify the application of a physically based snow energy balance model in a transitional snow climate and provide critical insight into precipitation phase sensitivity and snowmelt processes across time scales. We establish a foundation for improving and operationalizing this type of modeling within regular and flood forecasting workflows in the Northeastern United States.