A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Environment, Resources and Development

Thesis (M.Eng.) - Asian Institute of Technology, 1979

This study describes the development of a computer simulation model for daily mean potential evapo-transpiration (PET) on weekly basis. The daily potential evapotranspiration data was estimated, using the modified Penman equation, from 56-years of historical hydro meteorological data collected at a station located in the study area. The model was developed, primarily, based on the assumption that little seasonal variation occurs during a week. But during the course of the model development, it was found that the occurrence of an isolated rainy day or a sequence of rainy days during a week is important. However, the amount of rainfall has no significant effect on PET. One important outcome of this study is that the PET as a random variable for any week of year, is normally distributed and the weekly means of these distributions have a deterministic relationship. with the respective weeks of the year. The model makes use of these properties of PET in generating the data for any number of years. It is flexible enough to encompass the seasonal variations that occur during a week due to rainfall. For studying a complete crop production system, this PET model can be used as component of an over all crop environmental model describing the inputs of rainfall, evapo-transpiration and soil moisture variations with depth. A mathematical model of soil moisture storage and its variation with depth is also proposed in this study. The validity of this second model can be tested when the appropriate data required to tune the model becomes available.