WebGIS and Random Forest Model for Assessing the Impact of Landslides in Van Yen District, Yen Bai Province, Vietnam

Van Yen district in Yen Bai province of Vietnam is one of the most affected areas indicating high and very high potential of landslide occurrences. WebGIS technology is useful for dissemination of the geographical information related to landslides. To this aim, this paper provides a landslide information system for Van Yen district of the Yen Bai province. The paper firstly provides a landslide susceptibility map produced using a Random Forest-based model. Along the process, a landslide inventory from three investigations of 2013, 2017, and landslides along national road 32 and provincial road 163 (collected in November 2021) was considered. Additionally, thirteen factors were used in the model as variables including geological data, DEM (digital elevation model), slope, aspect, plan curvature, profile curvature, stream power index, topographic wetness index, fault network, river network, road network, land use-land cover data, and Sentinel-2 based NDVI (normalized difference vegetation index). The model was validated based on confusion matrix, and gave an excellent accuracy of 91.33%. Finally, WebGIS was created using open-source technologies such as Leaflet, Openlayers, Geoserver, PostgreSQL as the database management system, and PostGIS as it is plugin for spatial database management. WebGIS not only contains information relevant to landslides, but it also combines the landslide susceptibility map with population data in order to assess exposure for warning purposes.

Automated summary

This paper presents a landslide information system for Van Yen district in Yen Bai province of Vietnam, utilizing WebGIS technology for disseminating geographical information related to landslides. The paper develops a landslide susceptibility map using a Random Forest-based model and validates the model with a high accuracy of 91.33%. Additionally, a WebGIS is created using open-source technologies to combine the landslide susceptibility map with population data for assessing exposure for warning purposes.