Author(s): Roby Hambali; Djoko Legono; Rachmad Jayadi; Satoru Oishi
Linked Author(s): Satoru Oishi, Djoko Legono
Keywords: Short-term rainfall; Time series; Statistical properties; XMP-radar; Rain gauge
Abstract: Flood warning systems have become increasingly necessary to reduce hazard risk, especially lahar flow disaster. An effective lahar flow early warning system should provide information timely, thus allowing sufficient waiting time to respond. As an input to the lahar flow early warning systems, short-term rainfall information derived from a rainfall nowcasting model is needed. To select an appropriate nowcasting method, determining the patterns of historical data through its statistical properties is necessary. Some analyses of historical data can be used to find the trends that are further extrapolated forward. This study aims to analyze the statistical properties of short-term rainfall time series in the area of Mt. Merapi in order to find historical data patterns. The statistical properties of short-term rainfall time series were investigated through autocorrelation and spectral analyses. Several rainfall data with 2-minutes and 10-minutes time interval obtained from XMP Radar estimation have been adopted to determine the autocorrelation coefficient and spectral density. Collerogram was used to evaluate the characteristics of autocorrelation coefficients. The results show that there is high autocorrelation coefficient variability between locations within the radar coverage area, and no time lag between XMP radar rainfall and rain gauge rainfall for 10-minutes time-scale.
Year: 2018