Analysis of major orographic precipitation events over the Pacific Northwest using field data, mesoscale models, and an operational WSR-88D radar

The precipitation and microphysical structures are investigated for several major precipitation events over the Pacific Northwest using data obtained from recent field studies, such as the second Improvement of Microphysical Parameterization through Observational Verification Experiment (IMPROVE-2), the Penn-State NCAR MM5 and Weather Research and Forecasting (WRF) models, and two cool seasons worth of Portland, OR radar data. First, this presentation briefly summarizes results obtained for two heavy precipitation events during two IMPROVE-2 IOPs (4-5 December 2001 and 13-14 December), which were simulated down to 1.33-km grid spacing using the latest modifications to the Thompson bulk microphysical parameterization. The models realistically simulated the three-dimensional thermodynamic, kinematic, and precipitation structures as compared to aircraft data; however, the models tended to produce too much snow aloft. The surface precipitation overprediction was less for the 4-5 December event, since there was less riming, and much of the excessive snow aloft was advected into the lee.

In order to evaluate the model for a full spectrum of events, including more severe flooding events, the MM5 was integrated twice daily to hour 24 at 1.33-km grid spacing over southwest Washington and northwest Oregon for the 2005-2006 and 2006-2007 cool seasons using the latest Thompson scheme and GFS analyses for initial/boundary conditions. Portland, OR WSR-88D radar data were quality controlled, velocities dealiased, and data was interpolated to several 3D Cartesian grids with different spatial resolutions. Model u and v fields are used to compute radial velocities from the radar location to compare with the radar-observed radial velocities. Microphysical fields are compared in terms of mean values and frequency of occurrence above a threshold mixing ratio value. This presentation will highlight some of these comparisons between 3D volumes of model output and WSR-88D radar observations at Portland, OR, for some major precipitation events, which suggest that the MM5 is too robust with the snow cloud aloft as compared with the radar mean structures.

Finally, to evaluate the model during higher freezing level conditions (3-3.5 km ASL), the MM5 is verified during the major flooding event on 6-7 November 2006 over southwest Washington State using a combination of WSR-88D scanning radar data and a vertically-pointing radar.