Some Stochastic Models for Daily Preciptation Series 'Source Title' Modeling Hydrologic Processes, Proceedings Of 3rd International Hydrology Symposium
Bibliographic details
Summary
The occurrence of precipitation is directly related to the intensity of disturbance in the atmosphere. Atmospheric disturbance as well as precipitation are intermittent processes which makes the analysis and the generation of daily precipitation series very complicated. On the other hand, most of the meteorological factors causing those intermittent characteristics vary more gradually and continuously. Studies on the occurrence of precipitation have shown that the precipitation event seems to occur when n exceeds a threshold n, where n is an index represented by a linear function of atmospheric temperature, pressure, moisture, wind velocity, etc. The series consisting of precipitation event or no precipitation event, regardless of precipitation intensity, shall be called the wet and dry series. The Tetrachoric correlation analysis is used to analyze the spatial and temporal correlation structure of wet and dry series. the results of this analysis show that the autocorrelation of wet and dry series, i.e., the latent variation, is considerably high for some months and daily precipitation time series may not always be regarded as an independent or a weakly dependent time series. The writers also proposed to generate precipitation intensities by a scheme to generate approximately multivariate exponentially distributed random variables.