Seminar Abstract

Matt Brunengo

Thursday, October 11, 2007

A Monte Carlo model for simulation of rain-on-snow events in the Pacific Northwest

Western North America is susceptible to blasts of warm, heavy rain­fall in winter, when rain plus melting snow can cause high streamflows and erosion. In the Pacific Northwest, most major floods and landslides occur during such events. Recognition of these processes grew from the 1850s; research based on theory, instrumental records, and field sites elucidated many aspects of phenomena that became known as rain-on-snow (ROS). Geographic variability and sporadic occurrence complicate study of ROS, but some questions can be addressed through modeling.

A computer program combining probabilistic and deterministic elements performs Monte Carlo simulation of ROS events over hundreds of “years”, generating realizations of timing, initial and weather conditions. Within a model event, snow accumulates or melts, percolation is tracked, and water available for runoff at the ground is the output. Frequency distributions are based on data from the west-central Washington Cascades, some combined into curves and trends relating to elevation and date, so the model can be applied to specific sites or generalized elevations. One version calculates snow and percolation for measured weather conditions, to test algorithms and calibrate parameters. Validation is focused on Stampede Pass, site of a first-order weather station, snow course and SNOTEL installation.

The model is evaluated by comparing statistics and frequency–magnitude–duration relations of the instrumental record and model realizations, for precipitation, R+SM, and WAR. In the first application of the Monte Carlo model, the regional zone of greatest ROS enhancement to long-term WAR seems to lie at ~750 m; future uses include assessment of ROS in forests versus clearings, and estimation of possible consequences of climatic change.

Speaker bio:

Matt received B.S. (1976) and M.S. (1978) degrees in geology from Stanford University; and then pursued (but did not entirely catch) postgraduate studies in geology, geomorphology, hydrology and engineering geology at the University of Washington, in the geological sciences department and the Quaternary Research Center. After working on several consulting and project jobs he was hired by the Washington Department of Natural Resources, where he participated in studies of regional landforms, landslides, geologic hazards, and hillslope hydrology related to forest management activities and urban development. After the storm and floods of February 1996, Matt resuscitated his research on rain-on-snow, and enrolled at Portland State University to finish his doctoral research project. At the same time, he has worked in geological consulting, part-time for Washington DNR, and is currently teaching geology as a sabbatical replacement at Lewis & Clark College in Portland.