IIT project in India’s Mumbai uses Twitter to issue flood alerts | Floods News


When monsoon rains sweep into Mumbai annually, residents of India’s monetary hub discover their social media feeds awash with flood memes – from Venetian gondolas plying the town’s deluged streets to workplace staff commuting on inflatable dinghies.

This 12 months, a analysis institute hopes social media can play a extra sensible position – asking residents to tweet particulars about floods of their neighbourhoods and utilizing the information to concern geographically particular flood alerts in real-time.

“Since we can’t monitor flooding throughout the town ourselves, we considered taking the assistance of the neighborhood,” stated Subimal Ghosh, head of the local weather research division on the Indian Institute of Know-how Bombay, which is behind the initiative.

All over the world in nations together with Australia and Indonesia, local weather researchers are more and more utilizing information gathered through social media to assist monitor climate occasions equivalent to floods.

They are saying the information might be used to enhance emergency response and rescue efforts and make forecasting extra correct, which is able to turn out to be all of the extra very important as local weather change fuels excessive climate.

“We want participatory fashions for local weather adaptation as we might be hit by extra excessive rainfall occasions within the close to future. We will present real-time flooding like Google Maps reveals purple for visitors,” stated Ghosh.

India’s monetary capital, which lies on the coast of the Arabian Sea, has recorded “excessive” rainfall – outlined as greater than 200mm (7.9 inches) of precipitation in 24 hours – a couple of dozen occasions between 2017 and 2021, civic information reveals.

That’s double the variety of such occasions through the earlier five-year interval.

The Indian Institute of Know-how Bombay (IITB) mission, launched in June simply forward of the monsoon, will use digital elevation maps that present an space’s top above sea degree, spatial rainfall information, together with the tweets of residents.

Tweets noting “ankle-deep” or “knee-deep” water, for instance, might be harvested from Twitter utilizing an automatic system that may attempt to extract flood depth and site data from them.

This might be used to provide real-time flood data for the complete metropolis that might be put out on a portal.

As an example, “which areas want water to be pumped out first, how can the visitors be managed and the way can individuals attain dwelling?” Ghosh stated.

“The town wants a resilience plan,” he advised the Thomson Reuters Basis.

India monsoon
Folks cross flooded railway tracks subsequent to a parked passenger practice throughout heavy rains in Mumbai [File: Hemanshi Kamani/Reuters]

AI catastrophe bots

Comparable tasks elsewhere have proven how native individuals can play a component in crafting flood responses or alerts.

When flooding hit Australia’s east coast this 12 months, submerging total cities, Brisbane-based tech startup FloodMapp posted a collection of flood maps on its social media accounts, primarily Twitter.

The posts prompted some individuals to publish comparable maps of different areas. Others despatched images of the areas depicted, serving to to visualise what the maps had been exhibiting, a FloodMapp spokesperson stated.

“Folks have reached out to others to alert them to flood circumstances,” the spokesperson added in emailed feedback.

In a single case, a enterprise proprietor was capable of warn his workers to remain at dwelling after he noticed the map alongside a photograph from a neighbour, the corporate stated.

In Indonesia, an area web site based a couple of decade in the past, PetaBencana.id (which suggests catastrophe map), is utilizing synthetic intelligence (AI) and bots to map disasters in actual time.

The prize-winning platform asks individuals to confirm their social media posts with geotagged pictures after which combines official information to construct up-to-the-minute, on-line flood maps that “have been utilized by tens of millions of residents”, the platform’s founder, Nashin Mahtani, stated not too long ago.

South and East Asia are dwelling to almost 1.36 billion of the 1.46 billion individuals in danger from flooding all over the world, with India and China accounting for a 3rd, in line with the World Financial institution (PDF).

Low-income neighbourhoods are worse-affected by floods, significantly in cities like Mumbai which have a big inhabitants dwelling in casual settlements, the World Financial institution report states, pointing to poor drainage and land-use planning as key components.

Whereas the town launched a flood forecasting system two years in the past and put in huge drainage pumps in low-lying areas, storm-related disasters equivalent to constructing collapses and landslides are frequent within the monsoon months, researchers stated.

In current weeks, floods have hit Bangladesh and northeast India, killing greater than 25 individuals in India’s Assam state and leaving tens of millions stranded with little meals and consuming water in each nations.

“Flooding is an unavoidable actuality, and what we’re not capable of do is minimise the harm attributable to floods,” stated Prasoon Singh, convenor with Heart for International Setting Analysis of New Delhi-based The Vitality and Assets Institute (TERI).

“(However) crowdsourcing data can validate forecasts, assist tweak predictions and enhance reduction and rescue operations,” he stated, including that data garnered on social media may be fed into historic information units.

Within the quick time period, Ghosh stated he hoped the Mumbai mission would assist metropolis residents go about their every day enterprise – even through the monsoon season.

“My daughter was at school when it began raining closely and buses had been stranded on flooded roads. All of the mother and father had been anxious and we didn’t know tips on how to attain the varsity by means of waterlogged roads,” he stated.

“Actual-time data is essential.”

Leave a Reply