<![CDATA[Extreme weather can trigger a number of meteorological disasters such as landslides. This disaster may cause serious losses on livelihoods and hamper the growth of economy or development targets. This research focuses on determining a threshold triggering landslide events. Rainfall is considered as one of the common factor contributing to the landslide occurrence. The rainfall intensity as the trigger for landslides can be determined using a series of statistical methods. The threshold determination is performed using statistical techniques composed of Cumulative Rainfall Threshold (CT) and sorting analysis, dummy regression, cluster analysis, and change detection method. The methods are applied to determine the thresholds for landslides occurring in Sumatra Islands for the period of 2010-2017 retrieved from the website Data Informasi Bencana Indonesia managed by Badan Nasional Penanggulangan Bencana (DIBI BNPB). We evaluated daily rainfall data for the period of 2010-2017 compiled for 10 climate stations operated by Bureau Meteorology, Climatology, and Geophysics named in Bahasa Indonesia Badan Meteorologi, Klimatologi, dan Geofisika that are accessible for the Sumatra Island. The analyses suggest that the rainfall thresholds that should be monitored for detecting the potential occurrences of landslides in the study area are 15 mm per day, 30 mm per day, dan 65 mm per day. These values can be seen as warnings at different levels with the largest value, i.e., 65 mm, indicated higher confidence for the landslide event to occur. In other words, these values represent different levels of alert for the landslide occurrence that provide inputs for designing strategies of disaster prevention to mitigate the adverse impacts of landslide disaster.]]>
