Comparison of the Methods to Estimate Missing Values in Monthly Precipitation Data

Abstract

Estimation of missing data is essential in the meteorological, climatologically and hydrology analyses. This study employed the arithmetic mean method, normal ratio method, the modified normal ratio method, and correlation coefficient weighting method. The performance of these methods are then compared using correlation coefficient, the S-index, the root mean squared error and mean absolute error methods. The objective of this study is to determine the best estimation method for missing data for four precipitation stations in Makassar city. The results show that the modified usual ratio method is suitable to estimate the missing precipitation data in Makassar city. This study result could be useful information for climate research to complete the missing precipitation data, especially for rain gauge stations in Makassar city.