Comparison of Rainfall Interpolation Methods in Langat River Basin

Abstract

Rainfall is an element of climate that can be measured by a rain gauge. The rain gauge was set up for every station predefined by the Department of Irrigation and Drainage (DID) Malaysia. One millimeter (mm) of rainfall means that within a square meter of a flat surface, water can be as high as one mm. In the hydrology model, the rainfall data is very important in order to predict the flood or assist in the disaster mitigation plan. In this case, the availability of complete rainfall data in a region is essential. By performing spatial interpolation, rainfall data can predict values from the empty data at each point. In this study, Inverse Distance Weighting (IDW), Ordinary Kriging (OK), Simple Kriging (SK) and Kernel Smoothing (KS) method were considered in the rainfall interpolation for this area. Rainfall data at 20 points in Langat River Basin that obtained from DID Ampang for the period 2008–2017 were used as reference data. This study aimed to compare IDW, Kriging and Spline methods to obtain better interpolation methods. The interpolation is done by running a cross-validation using a geostatistical wizard in ArcGIS. The method effectiveness was evaluated by the calculation of mean error (ME), Root Mean Square Error (RMSE), Root mean Square Standardize Error (RMSSE) and Average Standard Error (ASE). For IDW method, only the ME and RMSE results are available. From the result, it can be seen that SK method outperforms the IDW, OK and KS method for these rainfall interpolations in Langat River Basin by showing better statistical evaluation.

Association of sociodemographic and environmental factors with spatial distribution of tuberculosis cases in Gombak, Selangor, Malaysia

“…Kriging is a proven method to interpolate since it is able to produce the lower errors and better predictions accuracy compared to other geostatistical interpolation methods such as Inverse Distance Weighting and Kernel Smoothing. Further, the fitted model in kriging does not only depend on the distance between the measured points and the prediction location but also the spatial relationships among the measured values around the prediction location [ 62 – 64 ]. Although with datasets consist of relatively few samples, kriging could produce low error estimations [ 65 ].…”

Prediction and Mapping Rainfall Classification Using Naive Bayes and Simple Kriging

“…Compared to previous studies [6], [7], [8], and [9] it only show prediction results, and does not display mapping. While research [10] and [11] only shows mapping that does not adjust the shape of the map. In comparison, this study produces predictive and mapping values that adjust the shape of the map.…”

“…S. S. Prasetyowati et al conducted research on air pollution in 2020 in the Bandung area and produced output in the form of a prediction map of air pollution in the next few years using the Simple Kriging method [10]. In the same year, the research of M. Hassim et al compared several interpolation methods to map rainfall in the Langat river area, Malaysia and the final result obtained is the simple kriging method is the most optimal method because it has the smallest RMSE value [11]. In 2021, [12] research will interpolate rainfall in Pakistan.…”