Author(s): Ratih Indri Hapsari, Satoru Oishi, Magfira Syarifuddin, Rosa Andrie Asmara
Linked Author(s): Ratih Indri Hapsari
Keywords: Debris flow, X-MP radar, uncertainty, critical rainfall, hazard
Abstract: Debris flow, which is the most serious secondary impact of volcanic disaster, is highly triggered by rainfall. Limited access to the area of active volcano slope as well as the risk of measurement gages damage due to disaster motivate to the use of X-band weather radars for observing the rainfall. However, the use of radar-rain prediction in sediment-related disaster involves high amount of uncertainties. In this study, debris flow disaster mitigation system by utilizing the high-resolution nowcasting products from X-band multiparameter compact (X-MP) and the assessment of its uncertainty are presented. The ensemble rain prediction by perturbing the advection vector is introduced aiming to gain the knowledge of uncertainties inherent in the system. The study area is the rivers on Mount Merapi that is historically the most active volcano in Indonesia. The benefits of ensemble forecasting over the single deterministic forecast are demonstrated, particularly to reduce missing of severe events. By evaluating five ensemble rainfall spatial distributions at once, the basin or village that is most likely to be hit by the debris flow is analyzed through hazard zoning. Rainfall critical line diagram is applied as a basis for judging the occurrence of debris flow. In order to evaluate the uncertainty of predicted rainfall temporal variation, snake lines development of hourly and working rainfall from the radar ensemble nowcasting products is drawn. The uncertainty assessment in the rainfall prediction system provides important information regarding the potential of disastrous debris flow. Verified with past occurrences, the integration of ensemble forecast product could provide a plausible range of the prediction possibility. This framework could be applied in disaster mitigation efforts for partially help to gage a confidence in the warning system
Year: 2017