洪天助 謝維權

USE of Statistical Data for Monthly Rainfall Forecast in Taiwan Area

Tien-Chu Hung Wei-Chyuan Shieh

Abstract

　　In order to forecast the monthly rainfall amount more accurately, the 30-year statistical data of monthly rainfall for some important stations in Taiwan and its vicinities were analyzed. The results show that there are several different types of rainfall pattern beyond the seasonal variations and the long period trends. Each of them was also discussed on the base of meteorological point of view.

　　In the second part of this article, three methods for rainfall prediction, i.e. the analogy method, the mean value method, and the ARIMA method, were presented, and the results were given and compared.