• Bergen, W. R., and Murphy A. H. , 1978: Potential economic and social value of short-range forecasts for Boulder windstorms. Bull. Amer. Meteor. Soc., 59 , 2944.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Blumen, W., 1985: Reflection of hydrostatic gravity waves in a stratified shear flow. Part I: Theory. J. Atmos. Sci., 42 , 22552263.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Brown, J. M., 1986: A decision tree for forecasting downslope windstorms in Colorado. Preprints, 11th Conf. on Weather Forecasting and Analysis, Kansas City, MO, Amer. Meteor. Soc., 83–88.

  • Cotton, W. R., Weaver J. F. , and Beitler B. A. , 1995: An unusual summertime downslope wind event in Fort Collins, Colorado, on 3 July 1993. Wea. Forecasting, 10 , 786797.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Doyle, J. D., and Coauthors, 2000: An intercomparison of model-predicted wave breaking for the 11 January 1972 Boulder windstorm. Mon. Wea. Rev., 128 , 901914.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Durran, D. R., 1986: Another look at downslope windstorms. Part I: The development of analogs to supercritical flow in an infinitely deep, continuously stratified fluid. J. Atmos. Sci., 43 , 25272543.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Durran, D. R., 1990: Mountain waves and downslope winds. Atmospheric Processes over Complex Terrain, Meteor. Monogr., No. 45, Amer. Meteor. Soc., 59–81.

    • Search Google Scholar
    • Export Citation
  • Haykin, S., 1999: Neural Networks: A Comprehensive Foundation. Pearson Education, 842 pp.

  • Insightful Corp., 2002: S-Plus. Version 6.1. University of Oklahoma.

  • Klemp, J. B., and Lilly D. K. , 1975: The dynamics of wave-induced downslope winds. J. Atmos. Sci., 32 , 320339.

  • Kretzschmar, R., Eckert P. , Cattani D. , and Eggimann F. , 2004: Neural network classifiers for local wind prediction. J. Appl. Meteor., 43 , 727738.

  • Laarhoven, van P. J. M., and Aarts E. H. L. , 1987: Simulated Annealing: Theory and Application. Reidel, 204 pp.

  • Lee, T. J., Pielke R. A. , Kessler R. C. , and Weaver J. , 1989: Influence of cold pools downstream of mountain barriers on downslope winds and flushing. Mon. Wea. Rev., 117 , 20412058.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Leptuch, P. A., 2001: Forecasting downslope windstorms in the vicinity of Boulder Colorado, an empirical statistical technique. M.S. thesis, School of Meteorology, University of Oklahoma, 200 pp. [Available from Bizzell Library, University of Oklahoma, Norman, OK 73019.].

  • Mathworks, 2002: Matlab. Version 6.5.0.180913a. University of Oklahoma.

  • Overland, J. E., and Bond N. A. , 1995: Observations and scale analysis of coastal wind jets. Mon. Wea. Rev., 123 , 29342941.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Richman, M. B., Santosa B. , and Trafalis T. B. , 2005: Feature selection of radar-derived tornado attributes with support vector machines. Preprints, Fourth Conf. on Artificial Intelligence Applications to Environmental Sciences, San Diego, CA, Amer. Meteor. Soc., J5.1. [Available online at http://ams.confex.com/ams/pdfpapers/87991.pdf.].

  • Rockwood, A. A., Brown J. M. , Weaver J. F. , Jamison B. D. , and Holmes R. , 1992: An expert system for the prediction of downslope windstorms. Proc. Fourth AES/CMOS Workshop on Operational Meteorology, Whistler, BC, Canada, AES/CMOS, 352–357.

    • Search Google Scholar
    • Export Citation
  • Sangster, W. E., 1977: An updated objective forecast technique for Colorado downslope winds. NOAA Tech. Memo. NWS CR-61, NWS Central Region Headquarters, Kansas City, MO, 24 pp.

  • Scheetz, V. R., Henz J. F. , and Maddox R. A. , 1976: Colorado severe downslope winds: a prediction technique. Final Rep., Contract 5-35421 to SDO/NWS/SDO/TDL from Geophysical Research and Development Corporation, 40 pp.

  • Scorer, R., 1949: Theory of waves in the lee of mountains. Quart. J. Roy. Meteor. Soc., 75 , 4156.

  • Smith, R. B., 1989: Hydrostatic airflow over mountains. Advances in Geophysics, Vol. 31, Academic Press, 1–41.

  • Tebaldi, C., Nychka D. , Brown B. G. , and Sharman B. , 2002: Flexible discriminant techniques for forecasting clear-air turbulence. Environmetrics, 13 , 859878.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Trafalis, T. B., Santosa B. , and Richman M. B. , 2003: Prediction of rainfall from WSR-88D radar using kernel-based methods. Int. J. Smart Eng. Syst. Design, 5 , 429438.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weaver, J. F., and Phillips R. S. , 1990: An expert system application for forecasting severe downslope winds at Fort Collins, Colorado, USA. Preprints, 16th Conf. Severe Local Storms, Kananaskis Park, AB, Canada, Amer. Meteor. Soc., 13–15.

  • Wilks, D. S., 1995: Statistical Methods in the Atmospheric Sciences. Academic Press, 467 pp.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 4 4 4
PDF Downloads 3 3 3

Statistical Modeling of Downslope Windstorms in Boulder, Colorado

View More View Less
  • 1 School of Meteorology, University of Oklahoma, Norman, Oklahoma
  • | 2 NOAA/Earth System Research Laboratory/Global Systems Division, Boulder, Colorado
Restricted access

Abstract

Downslope windstorms are of major concern to those living in and around Boulder, Colorado, often striking with little warning, occasionally bringing clear-air wind gusts of 35–50 m s−1 or higher, and producing widespread damage. Historically, numerical models used for forecasting these events had lower than desired accuracy. This observation provides the motivation to study the potential for improving windstorm forecasting through the use of linear and nonlinear statistical modeling techniques with a perfect prog approach. A 10-yr mountain-windstorm dataset and a set of 18 predictors are used to train and test the models. For the linear model, a stepwise regression is applied. It is difficult to determine which predictor is the most important, although significance testing suggests that 700-hPa flow is selected often. The nonlinear techniques employed, feedforward neural networks (NN) and support vector regression (SVR), do not filter out predictors as the former uses a hidden layer to account for the nonlinearities in the data, whereas the latter fits a kernel function to the data to optimize prediction. The models are evaluated using root-mean-square error (RMSE) and median residuals. The SVR model has the lowest forecast errors, consistently, and is not prone to creating outlier forecasts. Stepwise linear regression (LR) yielded results that were accurate to within an RMSE of 8 m s−1; whereas an NN had errors of 7–9 m s−1 and SVR had errors of 4–6 m s−1. For SVR, 85% of the forecasts predicted maximum wind gusts with an RMSE of less than 6 m s−1 and all forecasts predicted wind gusts with an RMSE of below 12 m s−1. The LR method performed slightly better in most evaluations than NNs; however, SVR was the optimal technique.

Corresponding author address: Andrew E. Mercer, School of Meteorology, University of Oklahoma, 120 David L. Boren Blvd., No. 5632, Norman, OK 73072. Email: amercer@rossby.metr.ou.edu

Abstract

Downslope windstorms are of major concern to those living in and around Boulder, Colorado, often striking with little warning, occasionally bringing clear-air wind gusts of 35–50 m s−1 or higher, and producing widespread damage. Historically, numerical models used for forecasting these events had lower than desired accuracy. This observation provides the motivation to study the potential for improving windstorm forecasting through the use of linear and nonlinear statistical modeling techniques with a perfect prog approach. A 10-yr mountain-windstorm dataset and a set of 18 predictors are used to train and test the models. For the linear model, a stepwise regression is applied. It is difficult to determine which predictor is the most important, although significance testing suggests that 700-hPa flow is selected often. The nonlinear techniques employed, feedforward neural networks (NN) and support vector regression (SVR), do not filter out predictors as the former uses a hidden layer to account for the nonlinearities in the data, whereas the latter fits a kernel function to the data to optimize prediction. The models are evaluated using root-mean-square error (RMSE) and median residuals. The SVR model has the lowest forecast errors, consistently, and is not prone to creating outlier forecasts. Stepwise linear regression (LR) yielded results that were accurate to within an RMSE of 8 m s−1; whereas an NN had errors of 7–9 m s−1 and SVR had errors of 4–6 m s−1. For SVR, 85% of the forecasts predicted maximum wind gusts with an RMSE of less than 6 m s−1 and all forecasts predicted wind gusts with an RMSE of below 12 m s−1. The LR method performed slightly better in most evaluations than NNs; however, SVR was the optimal technique.

Corresponding author address: Andrew E. Mercer, School of Meteorology, University of Oklahoma, 120 David L. Boren Blvd., No. 5632, Norman, OK 73072. Email: amercer@rossby.metr.ou.edu

Save