The technology gap in Wayanad landslides is emphasized by the horrific impact. This became obvious to Congress MP Shashi Tharoor, when he was talking about the way to predict landslides and their occurrence that is remote sensing technologies inadequate.
Pointing out that “we depend a lot on remote sensing. However according to experts this is not enough, and we need ground sensor grids for early warning of landslides. We don’t have it. And We have no ground sensors networked together. We have to collect more timely data.”
Various methods including deep learning are used in remote sensing to understand various aspects of an area such as topological, hydrological, vegetation among other geographical factors. Remote sensing satellites, which are specifically designed for this purpose survey and collect information about an area using photographs taken from space. These images show various attributes of the territory like water bodies, soils, plants and rocks among others in great detail. These images are then combined together to form some kind of map. The next phase involves incorporating machine learning algorithms into these maps to generate predictive models which can identify regions prone to specific ecological catastrophes such as earthquakes or floods etc., but as with most AI predictive models like AlphaFold (which creates 3D models of protein structures), there is no guarantee that they will be right.
According to Akshay Kumar who works at ERM as a geologist said “remote sensing does not predict anything because it only snapshots over time that may be studied”.
In contrast however Kumar indicates that ground radars are expensive ventures mostly done around mining sites or road construction activities. “Understandably certain hillsides should be considered off-limits since local authorities invariably lack funds for such expensive techs; however settlements lead to human lives loss.”
