The variability of rainfall rate will be examined over a range of time and space scales. The principal topics of the proposed research are the probability transformations of rainfall rate from the plot scale, at which physical process representations can be applied for land surface processes, to the observation scale of climatological remote sensing systems, and from the observation scale to the climate model scale. A statistical model of raindrop processes (Smith ?1991!) will be used to examine these transformations. The model of raindrop processes yields a stochastic/dynamic representation of rainfall rate and provides a unified framework for studying land surface processes related to rainfall rate and atmospheric processes related to remote sensing measurements of rainfall. In this study, empirical and theoretical analyses will be carried out for rainfall rate and raindrop processes. The lognormal cascade theory of Gupta and Waymire ?1990! will play an important role in both empirical and theoretical scaling analyses. Two principal problems will be examined: *How do distributional properties of rainfall rate and raindrop processes vary with scale? *How does rainfall rate depend on raindrop processes and how does this relationship vary with scale? An important element of this project is development of statistical procedures for assessing scale properties of rainfall and raindrop processes. The project will be an interdisciplinary effort involving hydrologists and statisticians. The proposed research will provide analytical tools for testing and refining the "probability distributed method" for representing land surface processes in climate models (Entekhabi and Eagleson ?1989!). Empirical studies will center on the southern plains regions of the U.S. and utilize radar data, rain gage data and surface drop-size data (including data from a network of disdrometers that will be deployed for the 1992 STORM field experiment). Surface drop-size data from a number of climatic regions in the U.S. will also be used for empirical studies.
