MULTIVARIATE SPACE-TIME ANALYSIS OF PRESTORM PRECIPITATION

This paper presents the methodologies and results of the multivariate modeling and two-dimensional spectral and correlation analysis of PRE- STORM rainfall gauge data. Estimated parameters of the models for the specific spatial averages clearly indicate the eastward and southeastw ard wave propagation of rainfall fluctuations. A relationship between the coefficients of the diffusion equation and the parameters of the s tochastic model of rainfall fluctuations is derived that leads directl y to the exclusive use of rainfall data to estimate advection speed (a bout 12 m s-1) as well as other coefficients of the diffusion equation of the corresponding fields. The statistical methodology developed he re can be used for confirmation of physical models by comparison of th e corresponding second-moment statistics of the observed and simulated data, for generating multiple samples of any size, for solving the in verse problem of the hydrodynamic equations, and for application in so me other areas of meteorological and climatological data analysis and modeling.