Home CATEGORIES Environment Uttarakhand Govt. Using Big Data to Prep for Natural Disasters

Uttarakhand Govt. Using Big Data to Prep for Natural Disasters

452
0
SHARE
 
Artificial intelligence, Internet of Things and Big Data are providing far-reaching solutions to the pressing issues facing humankind. Rapid digital innovation in these three technologies is delivering information through which governments, businesses, communities and citizens to make informed decisions and respond to crises. The state of Uttarakhand is harnessing geospatial Big Data in its disaster management plans in a similar manner.

Uttarakhand state harnessing Big Data

For those who thought Big Data processing and analysis is only restricted to corporations, it’s proving to be increasingly important in assessing natural hazards. Risk assessment in data-scarce locations, particularly when defining and scaling up the present human and economic value of assets, is another function it serves.
Last year, the Uttarakhand state government and the World Bank put together a team from software developer DHI Water & Environment, the Asian Institute of Technology (AIT) and the Evaluación de Riesgos Naturales (ERN) to complete a disaster risk assessment of the entire state. This was the first time the state government was proactively quantifying threat from natural hazards and the exposure of communities and critical infrastructure.
It was found that buildings are one of the major elements-at-risk. To overcome a gap of accurate information on the location of buildings, all building clusters and individual (distinctly standalone)
buildings, in Uttarakhand, were digitized from high-resolution satellite images covering the whole state. Considering the large number of buildings in Uttarakhand, the whole state was divided into 60,000 grids that were each randomly assigned to a data entry operator for digitizing using an application.
To be able to model the potential building losses, the team defined a set of typical building types (with consideration to varying vulnerability to hazards) and then estimated the proportion of each type in every settlement across the state. This was achieved by using a variety of Big Datasets and remotely-sensed data, such as topography and night-time light. These datasets were fed into a machine learning algorithm and then trained and validated using the results of detailed field surveys in representative villages and towns.

Big Data for tourists

Big Data also proved useful in modelling the spatial distribution of tourists and tourism activity around Uttarakhand. Not only is this good for the economy, tourists also benefit since they have limited risk knowledge of the places they are visiting.
Zones of high tourism activity were defined based on hundreds of thousands of anonymized spatial points, drawn from tourism booking and the review of websites listing hotels, restaurants and attractions, as well as photos posted on social media. Geolocating and characterizing these photos revealed popular hotspots, including informal infrastructures, such as tea-houses at highway viewpoints.

Big Data for flood hazard maps

Despite considerable efforts in modernizing its weather station networks, some parts of India do not currently possess adequate information from ground monitoring alone. In Uttarakhand, the team applied local weather data enhancement using alternative data sources, including satellite-based rainfall products (specifically, GPM and CHIRPS), together with other weather variables from the European Centre for Medium-Range Weather Forecasts (ECMWF) ERA-Interim climate models to produce final flood hazard and risk maps. In Uttarakhand, the team of experts statistically downscaled an ensemble of 14 climate projections to evaluate the impacts of climate change on local floods.
The multi-hazard risk maps developed for Uttarakhand are now being fed into a decision support system (DSS) for better disaster management in the state. The DSS integrates both baseline and real-time data to support the emergency operation centre at the state-level. For example, during the rainy season, the DSS will select the closest hazard and risk map in the at-risk areas and enable the authorities to plan the response activities for effective management of the disaster.