Predicting Cumulative Stream Power in Ungaged, Rural Watersheds

摘要

The vast majority locations where an assessment of historical stream power might be required (e.g., bridge sites) do not have the stream flow gage data needed for calculations. This paper describes and illustrates an approach to developing cumulative stream power estimates independent of time-series stream flow data, and using commonly-available geospatial and hydrologic data. This probabilistic method incorporates numerical watershed models and rainfall frequency estimates, coupled with a hydraulic model that translates flow rate into the depth and velocity. To illustrate, cumulative stream power was computed at a site (watershed area: 16 mi~2) according to two different methods: (1) utilizing available historical stream flow data, and (2) utilizing a watershed model coupled with probabilistic representations of rainfall depths. Findings show that in the shorter term (i.e., between 0-20 years), the difference between probabilistic (i.e., 'predicted') and gage-based (i.e., 'actual') cumulative stream power may differ considerably. However, when the probabilistic approach is applied over a sufficiently lengthy time period (i.e., beyond 20 years), the stream power computed by both methods can converge to less than 3% difference.