REGIONAL-SCALE WIND FIELD CLASSIFICATION EMPLOYING CLUSTER ANALYSIS
The classification of time-varying multivariate regional-scale wind fields at a specific location can assist event planning as well as consequence and risk analysis. Further, wind field classification involves data transformation and inference techniques that effectively characterize stochastic wind field variation. Such a classification scheme is potentially useful for addressing overall atmospheric transport uncertainty and meteorological parameter sensitivity issues. Different methods to classify wind fields over a location include the principal component analysis of wind data (e.g., Hardy and Walton, 1978) and the use of cluster analysis for wind data (e.g., Green et al., 1992; Kaufmann and Weber, 1996). The goal of this study is to use a clustering method to classify the winds of a gridded data set, i.e, from meteorological simulations generated by a forecast model.
- Research Organization:
- Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- W-7405-ENG-48
- OSTI ID:
- 15014321
- Report Number(s):
- UCRL-CONF-204808; TRN: US200803%%618
- Resource Relation:
- Conference: Presented at: 13th Joint Conference on the Applications of Air Pollution Meteorology with the Air and Waste Management Association, Vancouver, Canada, Aug 23 - Aug 28, 2004
- Country of Publication:
- United States
- Language:
- English
Similar Records
Mesoscale to Microscale Coupling for a Wind Ramp Case over Complex Terrain
Modeling the wind-fields of accidental releases with an operational regional forecast model