Analysis of orographic precipitation on Jeju-Island using regional frequency analysis and regression

The orographic effect is a common phenomenon in mountainous regions. Our goal is to analyze the orographic effect with quantile by regional frequency analysis and multiple regression. Multiple regression was used to develop models to estimate the amount and the spatial distribution of orographic precipitation in mountainous terrain using elevation, latitude, longitude, duration, and return period. Multiple linear regression analysis indicated that the model using the three parameters of elevation, latitude, and longitude, produces better results than four- or five-parameter models. Therefore, multiple nonlinear forms, the combination of the intensity-duration-frequency (IDF) relationship and the general linear regression form of orographic statistics were proposed to improve the accuracy of models. The models in this study showed an increase in accuracy of 18.31~86.27%. Moreover, these models produced good results in GIS analysis and were able to represent all cases examined in this study using only a few equations, in contrast to multiple linear models.