An Improved Temperature-Index Snowmelt Model on a Watershed Scale Using an Hourly Time Step for Real-Time Flood Forecasting in British Columbia, Canada (Poster)

One of the important indicators of a good real-time flood forecasting system is its time efficiency or the ability to produce timely forecasts. The majority of watersheds in British Columbia, Canada are large-scale snowmelt-dominated watersheds. Therefore, a simple but efficient snowmelt model helps improve time-efficiency of the real-time flood forecasting system. The temperature-index snowmelt model is one of this kind. However, it is more accurate when applied on a forest stand scale rather than on a watershed scale, and its time step is usually one day. The real-time flood forecasting system in British Columbia is the CLEVER Model, in which a watershed is simplified into a single node, and which adopts an hourly time step. In this study, the commonly used equation of temperature-index snowmelt model was improved and adapted to the application on a watershed scale using an hourly time step so that relative high forecasting accuracy and time efficiency were both achieved in the CLEVER Model.