HR: 0830h
AN: NG51A-0834    [PDF]
TI: On the Weibull Distribution as a Statistical Model for Surface Wind Timeseries in Reanalyses
AU: * Ong, E T
EM: eong@mcs.anl.gov
AF: Mathematics and Computer Science Division Argonne National Laboratory, 9700 S. Cass Avenue, Argonne, IL 60439
AU: Larson, J W
EM: larson@mcs.anl.gov
AF: Mathematics and Computer Science Division Argonne National Laboratory, 9700 S. Cass Avenue, Argonne, IL 60439
AB: The Weibull distribution has been identified as a statistical model for timeseries samples of surface station windspeed measurements. Previous studies have shown that near-surface winds from General Circulation Models can exhibit non-Weibull behavior. In this study we examine 10-meter windspeed data from the National Center for Environmental Prediction 50-year Reanalysis. We provide maximum-likeliehood estimates of Weibull distribution shape and scale parameters $\gamma$ and $\beta$ respectively, as well as estimates in the uncertainties in these parameters. We then use these estimates of $\gamma$ and $\beta$ with the Weibull distribution to create reference distribution functions for each timeseries sample and 1) estimate the mean and standard deviation and compare them with estimates computed from the sample, and 2) compare them with the sample using a one-sample Kolmogorov-Smirnov Test. Finally, we discuss possible reasons for non-Weibull behavior in the reanalyses, and describe how our results may be used in climate model validation.
DE: 3299 General or miscellaneous
DE: 3307 Boundary layer processes
DE: 3309 Climatology (1620)
DE: 3399 General or miscellaneous
SC: Nonlinear Geophysics [NG]
MN: 2003 Fall Meeting