Thursday, 15 January 2004
Partitioning of precipitation into rain and snow in distributed hydrologic simulations in the Western Cascades, Oregon, USA
Hall 4AB
One of the greatest challenges in hydrologic modeling in areas with significant orographic influences is accurate simulation of the precipitation fields, since this drives the streamflow response. In the northwest United States, where most of the precipitation occurs during the cool season, another major factor in streamflow simulation is the determination of whether the precipitation is falling as rain or snow, since these strongly influence the timing of the resulting runoff. The partitioning of precipitation in distributed hydrologic models into rain, snow, or a mixture of the two is often based on surface air temperature, since this is included in the station observation records that provide the precipitation and other meteorological data used to force the model. This study examines the adequacy for hydrologic modeling of using surface air temperature to determine this partitioning of rain and snow in the Santiam River basin, Oregon. The western slopes of the Cascade mountain range in Oregon, specifically an area including the South Fork of the Santiam River, was the geographical focus for second phase of the research effort dubbed Improvement of Microphysical PaRameterization through Observational Verification Experiment (IMPROVE-2). This intensive field observation campaign was carried out from 26 November through 22 December 2001, with measurements used to perform comprehensive verification of cloud and precipitation microphysical processes parameterized in mesoscale models. Included in the suite of IMPROVE-2 observations were both scanning and vertically pointing radar. While scanning radar observations in areas of complex terrain, such as the western Cascades, are problematic due to ground clutter and beam blocking, vertically pointing radar does not suffer from this. We show that, by replacing the surface air temperature-based algorithm in a distributed hydrologic model with a freezing level determined with S-band radar supplemented by other observations, significant improvement in the simulated hydrograph can be obtained.
Supplementary URL: