Group Page

Sponsored by the Northwest Modeling Consortium, qwe have run the MM5 mesoscale model operationally at high resolution since 1995. Currently, the MM5 is run at 36 km horizontal resolution over the eastern Pacific, 12 km over the Pacific Northwest, and 4 km over Washington, Oregon, and Idaho. We are also running and evaluating the WRF model. The thrust of our work is to determine the benefits of high-resolution forecasting, improve verification approaches, and to identify and address weaknesses in model physics. We are now evaluating various approaches to grid-based bias removal to take out systematic biases in surface fields.

Improving Microphysics and Precipitation Parameterizations in Mesoscale Models

Working with the UW Cloud/Aerosol Physics and Mesoscale Groups, we completed a two-phase field experiment (IMPROVE) that provided comprehensive microphysical and basic state data for improving moist parameterizations in mesoscale models. We are currently analyzing this field data and comparing them to high-resolution model simulations to determine the problems with moist parameterizations and are working on various parameterization improvements.

The Mesoscale Meteorology of the West Coast of North America

We are studying a number of weather phenomena of the west coast of North America. One major project has examined the structure and dynamics of the gap winds in the Columbia River Gorge with both observational data and MM5 modeling.

Major Forecast Failures and the Evaluation of Pacific Initialization and Its Effects on Forecast Skill

Using all available observational resources, we are evaluating initialization quality over the Pacific of major forecast models, and the effects of initialization quality on forecast skill. In concert, with Lynn McMurdie we are examining the origin of major forecast failures over the eastern Pacific.

Use of Ensembles for Mesoscale Numerical Weather Prediction

We are testing the usefulness of running a high-resolution ensemble predictions using a number of operational initializations and varying physics to test this approach to statistical numerical weather prediction. The UW Ensemble system is one of the highest resolution (12-km) ensemble systems in the world and we are examining its ability to provide calibrated probabilities and to predict forecast skill. Currently, there are 17 members run twice a day. As part of this project we are working closely with the Statistics department (Adrian Raftery and Tilmann Gneiting) and others in the DOD-sponsored UW MURI effort on post-processing of the ensemble forecasts. We are also working with Professor Greg Hakim's group on the development of an Ensemble Kalman Filter system for data assimilation and ensemble forecasting. A major new project is to use ensemble and 4D-Var approaches for mesoscale data assimilation.

Regional Climate Simulations

In concert with Eric Salathe of the Climate Impacts Group, we are now running the MM5 for ten-year periods (2070-2080, 2045-2055, 2020-2030, 1990-2000) at 45 and 15 km resolution using forcing from the PCM and ECHAM global circulation models to determine the regional implications of global warming. We hope to learn whether the land-water contrasts and terrain of the region might produce any unexpected regional implications for global climate change.
* Northwest Windstorms
This project is studying the nature of the intense Northwest windstorms associated with deep, synoptic lows that often pass through the Pacific Northwest during winter. Some of these storms have had winds exceeding 125 mph, with the greatest equivalent to Category III hurricanes in wind speed and damage potential.

Application of Meteorological Data For Regional Transportation Needs

In a cooperative effort with the Washington State Department of Transportation, real-time observations and high-resolution model output are being made available to Washington State travelers. This work includes web-pages tailored to road maintenance personnel and the traveling public.

Cooperative Research and Interaction with the National Weather Service

With support from the NWS CSTAR program, we are working to develop mesoscale analysis and prediction tools for operational application. These include model grid-based bias removal and ensemble Kalman filter data assimilation. We also work closely on regional prediction and co-sponsor the Northwest Weather Workshop.

Dissemination of Weather Information to the Public

As part of this effort, the PI provides a weekly weather forecast and discussion on KUOW at 10:52 AM on Friday, in addition to regular public lectures at schools, fraternal organizations, and other groups.