Development of a Community-Based Anticipatory Action Protocol for Landslide Early Warning Systems

Abstract

Landslides pose significant risks to communities in Jombang Regency, East Java, particularly under prolonged rainfall. This study developed an operational anticipatory action protocol that integrates an Internet of Things (IoT)-based landslide early warning system with community observations and multi-channel risk communication in Jumok Hamlet, Sambirejo Village. The study used a community-based approach through a focus group discussion (FGD) with 30 purposively selected participants, field and geological surveys, protocol development, and qualitative validation by experts and community users. The protocol was structured around community-based early warning principles and translated hazard information into four graduated levels: Normal, Advisory, Watch, and Warning. The resulting protocol combines daily rainfall above 61 mm, three-day cumulative rainfall above 91 mm, ground movement above 2 cm/day, progressive crack widening, and soil saturation as operational indicators. The FGD identified widening ground cracks, prolonged rainfall, increasing soil saturation, unusual ground movement, and standing water during soil cultivation as locally recognized warning signs. Expert validation focused on information flow, operational procedures, and stakeholder coordination; user validation focused on SOP readability and usability. Feedback led to clearer trigger-to-action sequences and simpler, context-specific wording. The protocol has not yet been tested during actual landslide events. It links hazard indicators with predefined communication channels and anticipatory actions, including drainage management, mobilization of vulnerable groups, evacuation preparation, and self-evacuation. The study demonstrates how technological monitoring and locally observed indicators can be combined to translate warnings into coordinated community action.