The Sub-Daily Distribution of Snowmelt

Abstract:

The hydrologic cycle in many mountainous headwaters around the world have snowmelt dominated hydrographs. In addition to water resources for communities and ecosystems, high rates of snowmelt can cause flooding that results in damages to infrastructure. The standard in the United States flood forecasting looks primarily at rainfall estimates but lacks estimates for high rates of snowmelt in regions such as the Southern Rocky Mountains. Recent studies have shown that events such as a 10 year 24 hour snowmelt event is as much as 45% greater than the same recurrence interval rain event. Additionally, this 24 hour snowmelt likely occurs over a much shorter time period due to snowmelt being primarily driven by solar radiation. This study presents and tests a sub-daily temporal distribution of snowmelt. The snowmelt distribution presented herein is tested against hourly data for known daily melt rates from snow telemetry (SNOTEL) stations, and then for conditions when weekly or bi-weekly snow loss is known. It is additionally utilized for modeling a one-dimensional soil profile for infiltration across the soil-snow interface. The intent of this study is to create a less computationally intensive method than the mass energy approach and improve upon the simple degree-day method for the representation of snowmelt at sub-daily time steps. This can be used for streamflow, groundwater recharge, soil moisture distribution, and other land surface modeling efforts. Results of the study display strong agreement with hourly SNOTEL data from Colorado Front Range stations for an assumed 8-hour melt period. Peak flow estimates from snowmelt driven floods could be estimated from long-term datasets to calculate frequency of these flood events. Further application of this sub-daily distribution of snowmelt could be for partially or fully glaciated watersheds with modifications for differences in latitude and/or elevation causing longer or shorter periods of melt per day.