Water Resources Research Act Program

Details for Project ID CO_2025_SMALL

Improving regression-based streamflow forecasting in the Upper Colorado River Basin using machine learning clustering

Institute: Colorado
Year Established: 2025 Start Date: 2026-01-01 End Date: 2026-12-31
Total Federal Funds: $10,000 Total Non-Federal Funds: $10,000

Principal Investigators: Eric Small

Project Summary: Snowmelt provides approximately 80% of Colorado's water supply, making accurate monitoring of melt timing critical for water resource management. However, current in-situ methods of measuring snowmelt onset are not suitable. Synthetic Aperture Radar (SAR), specifically C-band (~5cm) from Sentinel-1, shows promise in measuring liquid water presence in a snowpack, which is a proxy for determining snowmelt onset. But SAR faces challenges in high-incidence angles and densely vegetated areas. This research proposes a deep learning (DL) model that leverages C-band SAR measurements to extend snowmelt onset mapping into complex terrain by detecting liquid water presence in the snowpack. The DL model will be trained and validated using field-collected data to create a near-real-time (NRT) snowmelt onset mapping product for the South Platte and provide a framework for applying this methodology to all of Colorado's watersheds.