ARS | Publication request: THE DEVELOPMENT AND TESTING OF A SPATIALLY - RELEVANT STOCHOSTIC WEATHER SIMULATION MODEL

Submitted to: Climate and Weather Research Workshop Proceedings
Publication Type: Proceedings
Publication Acceptance Date: September 15, 1995
Publication Date: N/A

Interpretive Summary: There typically is no readily available solution for the situation when climate data are needed for a particular application at a location but no such data exist. As a possible solution to this dilemma, a methodology has been developed and is now being tested on a two-state regional scale whereby the necessary information needed to operate a statistical weather generator model is determined for any given location in a specified domain. A new spatial distribution model called PRISM is linked to the USCLIMATE weather generator model, enabling weather sequences of any length period to be generated at any location. The paper describes the development of this method, and its application over parts of Idaho and Oregon, U.S.A.

Technical Abstract: A methodology for spatially distributing the basic parameters of a stochastic weather simulation model is developed and applied over a specific region with a variety of topographic regions and climates. Development of baseline parameters for climate stations with relatively short periods of record is described. The stochastic climate model, or weather generator, USCLIMATE, was chosen for parameterization, and the spatial interpolation program PRISM was chosen for spatial distribution of parameters. Initial testing of this procedure is for maximum and minimum temperature and precipitation, over a regional domain of approximately 60,000 km2.