This article develops a localized landslide early warning system (EWS) for Rohingya refugees and host communities in Cox’s Bazar district in Bangladesh. The arrival of nearly 750,000 Rohingya refugees in Bangladesh since 2017 has led to extensive deforestation and hill cutting in Cox’s Bazar district. Refugee camps are located on hills and loose soil, making them highly vulnerable to rainfall-triggered landslides.
The authors employ advanced geoinformation techniques to develop the landslide early warning system (EWS), integrating landslide susceptibility zones, rainfall thresholds, and forecasted daily rainfall data. The analysis relies on land cover maps for four years (1998, 2001, 2017, and 2018) prepared using images from the Landsat satellite missions. The authors produced landslide inventory and factor maps using historical landslide information, from which they constructed landslide susceptibility maps, categorized into three susceptibility zones (low, medium, and high). Rainfall thresholds for different susceptibility zones were analyzed using historical rainfall information following landslide events.
The landslide EWS takes into consideration land cover changes, historical landslide events, local rainfall thresholds, landslide susceptibility maps, and a hazard matrix to dynamically relate 5-day forecasted rainfall and their spatial association with the susceptibility map.
Main findings:
- Approximately 5,800 hectares of forest land cover disappeared due to the 2017 Rohingya influx. Grassland type that contains deciduous forests was significantly reduced because of the influx.
- Land cover changes through hill cutting and slope modifications, and unplanned urbanization are predominantly responsible for slope failures.
- Consecutive 5-day periods of rainfall between 95–220 mm could initiate landslides in high susceptible areas.
- The EWS effectively predicted the 10 September 2019 landslide event, which was triggered by 422 mm of rainfall in 24 hours.
The authors conclude that a large area has been deforested to build the refugee camps in Cox’s Bazar district that has significantly increased landslide vulnerability of the Rohingya refugees and their host communities. The authors propose early warning scenarios for low, medium, and high susceptible zones that would be triggered by forecasted rainfall exceeding specific thresholds. Such an early warning system would support local authorities and international organizations to reduce disaster risks and save lives from landslides in Cox’s Bazar district. Moreover, the proposed EWS is replicable and can be contextualized in similar settings.