Spatial Rainfall Rate Estimation from Multi-Source Data in Klang Valley, Malaysia

Abstract

This study investigates rainfall distribution by estimating rainfall rate from multi-source data of rain gauge, radar image and TRMM in Klang Valley, Malaysia. The study is also looking into rainfall intensity to identify rainfall types based on 35mm/hr thresholds for 5-minute interval using ground rain gauge data measurement.  The results revealed that during the study period, the stratiform rainfall type was found dominant, and most of the rainfall events occurred during the evening. The simple regression and bias error analysis have conducted to assess the potential of radar image and TRMM for rainfall estimation rate. It has shown a positive but relatively weak relationship of regression coefficient between the rain gauge measurements and both data sources. The results indicated that radar image has better performance than TRMM satellite in rainfall rate estimation over Klang Valley. The radar has estimated a total rainfall rate of about 42.5mm/hour with percent bias are (-14.49%) of error relative to rain gauge data measurement. Meanwhile, the per cent bias for TRMM tends to underestimate the rainfall measurement by (-42.05%) with only 28.8mm/hour total rainfall rate were estimated. The spatial interpolation of the IDW technique reveals the rainfall distribution pattern in the study area, interpolated rainfall distribution from a radar image has shown a good agreement with rainfall distribution from rain gauge data measurement. Although radar image has higher accuracy in rainfall rate, estimation due to limited data availability used in this study was unable to reveal the rainfall pattern. Furthermore, both data products used in this study show a lower ability to detect high-intensity rainfall events due to limited data availability appropriately.