Keith A. Brewster, Ph.D.

Senior Research Scientist
Adjunct Associate Professor

Center for Analysis and Prediction of Storms
University of Oklahoma
120 David L. Boren Blvd., Suite 2500
Norman, OK 73072

Phone: (405) 325-6115
Fax: (405) 325-7614
Office: National Weather Center, Room 4238
e-mail: kbrewster@ou.edu

Curriculum Vita (pdf)

Publications List (pdf)

Dissertation Abstract





Phase Correction Developed for Thunderstorm Forecasting

It is well known that it is common for numerical forecasts to have phase or position errors. Total forecast RMS errors can be large if thunderstorms or other features such as fronts are mis-positioned. Radar and mesonet data can tell us a lot about the position of such features, but are incomplete in terms of describing the complete temperature, wind and hydrometeor distribution within the features.

I have been developing techniques to directly deal with such phase errors in the analysis and assimilation of mesoscale and storm-scale (grid spacing of 10-km or less). Tests have been performed on frontal analysis using mesonet data and on the initial development phase of thunderstorms in a severe weather outbreak in the Texas panhandle during VORTEX-95 (June 8, 1995). Results are detailed in my Ph.D. dissertation and have been conditionally accepted by Monthly Weather Review.

Phase Correcting Data Assimilation: Part I

Article Text (206 kB pdf)
1st set of figures (618 kB pdf)
2nd set of figures (568 kB pdf)

Phase Correcting Data Assimilation: Part II

Article Text (194 kB pdf)
1st set of figures (571 kB pdf)
2nd set of figures (737 kB pdf)
3rd set of figures (827 kB pdf)
4th set of figures (95 kB pdf)




ARPS Data Assimilation System, ADAS

My current work focuses on several aspects of numerical forecasting of thunderstorms and their environment. I am especially interested in improving forecasts of thunderstorm initiation. As part of the team working on the The Advanced Regional Prediction System (ARPS) , I am investigating improving the real-time forecasts made during VORTEX in the Spring of 1995 and again in the Spring of 1996 as an independent project. To make a forecast you must have a good mesoscale analysis. Toward that end, I have developed the ARPS Data Analysis System, ADAS . ADAS is based on the Bratseth analysis method, an iterative approximation to the statistical method known as optimal interpolation. This implementation uses the NCEP Rapid Update Cycle model as a background. Data from the Oklahoma mesonet, surface aviation observations, network radiosondes and wind profilers are included in the products available on the World Web Web. Separately, we are testing the ingest of Doppler radar wind and reflectivity information, using data from NIDS and level-II WSR-88D data from the Twin Lakes (Oklahoma City) radar. This has been installed in ADAS and will incorporated into a full assimilation system real soon now.





ADAS Installed in Korea.

During a recent visit to Korea, working with the staff of the Korean Meteorological Administration (KMA) we were able to set up an ingest system for ADAS to use the 400-station Korean Automated Weather Station meso-network and the sounding and surface data available on GTS. This will be used in a real-time test of ARPS during their 1997 monsoon season, known locally as changma. For 1998, data from the KMA radar network will be ingested.





ADAS Application in Utah

ADAS, which is released as part of the ARPS software package, has been applied in many domains. For example, at the University of Utah ADAS is being run in real-time using some local mesonet data in northern Utah. Researchers John Horel, Steve Lazarus and Carol Ciliberti have added their own enhancements to ADAS based on their experience in Utah.



n0iaw QSL card

Storm Spotting

As a severe storm spotter and semi-retired storm chaser, I keep up with storm spotting activities in the Great Plains. I maintain a list of spotting frequencies a guide to TV storm coverage, and documentation on the NWS WSR-88D, Doppler radar network. A must-read for any aspiring chaser is the safety spiel by Dr. Chuck Doswell of the National Severe Storms Lab.



Other Interests