Comparison of stochastic models of rainfall processes in mountainous climatic conditions

Abstract

The main objective of this work was to find a stochastic simulation model that was suitable for designing a rainwater harvesting system for agricultural water utilization and irrigation water in a mountainous region. Several models were applied using daily rainfall data from two sites (Gangneung and Daekwanreung) to assess the accuracy and suitability of the model for simulating the daily rainfall. The amount of rainfall for the mountainous region was well described by a two-parameter gamma distribution and performed better than other distributions. However, validation tests revealed that the annual mean absolute percentage error (MAPE) was more than 10% at both locations. This result was different from some previous research in which a three-parameter mixed-exponential distribution was well described. In addition, although the exponential distribution was also well described by a second-order expression, the use of a one-parameter distribution had a simpler algorithm than the two-parameter gamma distribution and the three-parameter mixed-exponential distribution.