INTEGRATION OF GEOSPATIAL BASED CONCEPTUAL APPROACH IN A LANDSLIDE WARNING SYSTEM

Abstract

A disaster occurs when an extreme geological, meteorological, hydrological or any intolerable human event exceeds the ability of a community to cope with that event. One such disaster is landslides. It is of the utmost importance to understand its causes and to generate timely warnings. For this, a warning system needs to be constructed. From a comprehensive vulnerability assessment and hazard mitigation standpoint, it is necessary to look beyond the effects of an event on a single system as it is quite complex. Warning system will require more than just improved engineering and technology. A landslide warning system requires large amount of spatial and non-spatial data and disaster modelling/ predicting techniques, apart from communication techniques. Spatial information can be derived from medium and high resolution satellite images. Sequential images can provide information on changes in landscape. The spatial information along with non-spatial data can be effectively integrated in GIS for modelling the natural disaster phenomenon which may be ultimately used predicting landslides. Similar approach has been adopted here as well; a landslide prediction tool is developed using Arc Objects, VBA in Arc GIS for Uttarakhand region. Here as you enter all the input parameters, the model automatically performs susceptibility analysis and landslide prediction is done. Some of the parameters have also been made dynamic in this model so that susceptibility calculation can be done in future with revised parameters. Analyzing conditions existing in the landslide predicted area, a mail is sent to local authority for taking further action. This mail contains all important information regarding the landslide predicted area. This model is an initiative towards landslide prediction, such models are really the need of today for making effective warning. Also the warning generated by model should be quickly disseminated to people in real need of it. The probability calculated should be highly accurate, as one false warning could really break the confidence of people on warning.