Developing a Landslide Early Warning System for Chittagong City, Bangladesh
Abstract
The study aims to develop a Web-GIS based landslide early warning system (EWS) as in recent years, landslides in Chittagong Metropolitan Area (CMA), Bangladesh caused catastrophic damages to lives and properties. The triggering factors were torrential rainfall in short duration and people living on dangerous slopes by cutting hills and destroying forests. Eleven factor maps - soil permeability, surface geology, landcover, altitude, slope, aspect; and distance to stream, fault line, hill cut, road cut, and drainage network - along with a detailed landslide inventory map were produced. Artificial neural network (ANN), multiple regressions, principal component analysis and support vector machine methods were applied to produce landslide susceptibility maps. After model validation, the best-fit ANN map was classified as never warning, low, medium, and high susceptibility zones. Rainfall threshold analysis (1960-2017) reveals consecutive 5-day rainfall between 70-160 mm could initiate landslides in high susceptibility zone. The EWS was developed using various libraries and frameworks - MySQL, PHP, JavaScript, OpenLayers, W2UI, and jQuery along HTML5 and CSS. The system is publicly available, dynamic, replicable in a similar context and able to disseminate landslide warnings four days in advance via email notifications. The EWS is original, first of its kind in Bangladesh and expected to reduce landslide disaster risks.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFMNH13A..05A
- Keywords:
-
- 0540 Image processing;
- COMPUTATIONAL GEOPHYSICSDE: 4315 Monitoring;
- forecasting;
- prediction;
- NATURAL HAZARDSDE: 4333 Disaster risk analysis and assessment;
- NATURAL HAZARDSDE: 4339 Disaster mitigation;
- NATURAL HAZARDS