Close Window

Univ. of Oklahoma: "Forecasting downslope windstorms"

Final Report


The purpose of this project was to develop an objective method of predicting peak wind gusts in Boulder, CO, a location prone to periods of severe (60 mph or greater) downslope wind gusts. Our prediction method is based on identifying conditions upstream from the verification site that are favorable to the occurrence of downslope windstorms at the verification site. The time it takes for the upstream parameters to be advected to the verification site in Boulder is the predictive lead time. The prediction method consists of calculating twenty quantities, or predictors, from upstream rawinsonde soundings, which attempt to quantify the upstream momentum and stability characteristics. These predictors are inserted into a statistical model that predicts a peak wind gust for a trailing verification period. The parameters range from simple ones such as the 700 mb geostrophic wind speed and direction, the 3,700 m normal-to-barrier wind component, and the 700 mb geostrophic temperature advection, to theoretically derived predictors, which attempt to quantify the intensity of lower-tropospheric stable layers and the tilt of lee waves. J. M. Brown at FSL and E. Thaler at NWS supplied the verification data. P. Leptuch, with advice from H. Bluestein and M. Richman at OU and from J. M. Brown at FSL, developed the statistical models. The statistical model software is being made available to the NWS/Boulder. P. Leptuch will be awarded an M. S. degree in meteorology from OU in May 2001 for his efforts. Using the perfect prog approach, the actual NWS application will use predicted soundings, interpolated to sounding locations used to derive predictors identified as useful by this study. These predicted soundings (most likely, from NCEP model output available in BUFR format) will then be used as input to the statistical engine derived as part of this study.




Brown, J. M., 1998: Thoughts on the use of NCEP model output as mesoscale forecast guidance in the West. NWS Western Region Great Divide Workshop, Great Falls, MT, 23-25 Sept. 1998.

Brown, J. M., 1998: Interpretation of mountain waves in the RUC and Eta models: Brief review of forecasting considerations for Boulder windstorms. NWS Denver Winter Weather Workshop, Oct. 1998.

Brown, J. M., 1999: A review of the severe windstorm of 2-3 Feb. 1999. Informal seminar, NOAA/Climate Diagnostics Center, Boulder, CO, 5 Feb. 1999.

Leptuch, P., J. M. Brown, H. B. Bluestein, E. Thaler, and M. B. Richman, 2000: Forecasting downslope windstorms at Boulder, Colorado: The empirical-statistical approach revisited. Preprints, 9th Conf. on Mountain Meteor., Aspen, CO, Amer. Meteor. Soc., 105-108.

Leptuch, P. A., 2001: Forecasting Downslope Windstorms in the Vicinity of Boulder, Colorado: An Empirical Statistical Technique. M. S. Thesis, School of Meteorology, Univ. of Okla., Norman, 101 pp (final draft not yet completed as of time of final report).

Brown, J. M., 2001: Downslope windstorms and related phenomena. Presented at MSC/COMET Winter Weather Course, 20 Feb. 2001.

Brown, J. M., 2001: Thoughts on forecasting high wind over SE Wyoming. To be presented at Mountain Weather Workshop, Cheyenne, WY (sponsored by WFO

Leptuch, P., J. M. Brown, H. B. Bluestein, E. Thaler, and M. B. Richman, 2001: Forecasting downslope windstorms at Boulder, Colorado: An empirical-statistical approach. Wea. Forecasting. (in preparation, to be submitted by summer 2001)


4.1 The University of Oklahoma benefited substantially from the collaboration with J. M. Brown at NOAA/FSL and E. Thaler at NWS/Boulder. P. Leptuch learned about the processing and analysis of real meteorological data and how to apply statistical models to help solve forecasting problems. H. Bluestein developed and taught a section on downslope windstorms, in the fall of 2000, in his graduate course Advanced Mesoscale Meteorology. J. M. Brown was instrumental in providing the dynamical background and E. Thaler was instrumental in providing the forecasting motivation. Mr. Leptuchs progress was slowed down because he studied for the Ph. D. qualifying exam during the summer of 2000 while he was also preparing his presentation at the Mountain Meteorology Conference. In the spring of 2000 and 1999 he also participated in my Severe Storm Intercept Project. However, following the exam, conference, and field program, he was able to complete the project.

4.2 The NWS forecast office in Boulder will have a new tool to aid in forecasting peak wind gusts in future wind storm seasons. The software that P. Leptuch has developed will be incorporated into forecast operations via the AWIPS system by the 2001-2002 windstorm season. This statistical approach will be combined with output from mesoscale forecast models to aid in the issuance of high wind watches and warnings. The severe windstorm forecast problem still remains one the most difficult issues faced at this forecast office and it looks forward to having another tool to aid in attacking this problem.