Multiple Imputation for Hydrological Missing Data by Using a Regression Method ( Klang River Basin )

Rainfall amounts and water surface elevation are considered as one of the most important climatic parameters. Because these two parameters will have a direct impact on water resources management decisions such as meet the water needs and prevent flooding. But in some cases, for some reason all time series data are not fully recorded. To fill the gaps in the data, several interpolation methods currently used. One of these methods is regression analysis as a statistical method. By using regression, we can determine the mathematical relationship coefficients between inputs and outputs. By achieving the equation, we can obtain the unknown quantities. In this research, the daily data between 2005 to 2015 for 5 Rain-gauge stations and 3 elevation measurement of water surface stations in the Klang River Basin were used. The main goal was to find the missing value of the water level in the mentioned three stations by rainfall and water level data. To evaluate the obtained results, Multiple R, R2, and Standard Error were used. The results indicate that the standard error in normalized data was less than the regular data. Multiple r values for the Klang at Taman Sri Muda1, Klang at Jam, Sulaiman, WP and Klang at Emp Genting Klang, WP are 0.35, 0.42 and 0.28, respectively.

Filling gaps, Interpolation, Data mining, Statistical method, Rainfall, Water level