Estimation of precipitation data by using deterministic interpolation methods: A case study in Penang Island

Abstract

Heavy rainfall events can make slopes become unstable and consequently lead to terrestrial landslide which caused damage to properties and loss of life. However, the precipitation data of high risk landslide hazard area cannot be obtained due to difficulty in placement of rain gauge at mountainous or hilly areas. Hence, the aim of this study is to find out which deterministic method is suitable for estimating the precipitation value of landslide hazard area in Penang Island. Precipitation data of rainfall stations at Penang Island are secondary data that was collected from Department of Irrigation and Drainage (JPS). Four different deterministic methods include old normal ratio method (ONRM), inverse distance weighting method (IDW), modified normal ratio based on square root distance method (MNRT) and modified old normal ratio with inverse distance method (ONRIDW) are studied to estimate the precipitation data of the target station. It is found that the target station better locates at the center of the surrounding stations. Besides, the optimal distances from the target station to the surrounding stations are within radius of 10 kilometres. The performance of all the estimation methods are investigated using mean absolute error (MAE), similarity index (S-index) and coefficient of correlation (R). It is found that IDW with the power parameter p equals to 1 and ONRIDW are found to perform better on precipitation estimation in Penang Island. Hence, IDW with p equals to 1 and ONRIDW can be used for estimating precipitation data of high risk landslide in mountainous area such as in Penang Island.