Mumbai, like many other cities in India, experiences frequent extreme rainfall events resulting in severe floods and waterlogging.
However, there is currently no system in place to monitor flood situations in real time, leading to difficulties in managing and responding to flooding.
Using Twitter as a Crowdsourcing Tool
A team of researchers led by IIT Bombay has found a way to monitor real-time flooding by crowdsourcing data from Twitter.
They developed an automated program to retrieve flood-related data from tweets, which includes water logging levels and location information.
This approach enables them to provide real-time information about flooding in Mumbai.
Study Details
The research findings, yet to be peer-reviewed, have been posted on the arXiv preprint server.
Dr. Subimal Ghosh and his team from the Department of Civil Engineering at IIT Bombay led the study.
They evaluated the usefulness of crowdsourced flood data by analyzing data from an extreme rainfall event on July 5, 2022.
Historical flood-related tweets were also used to generate flood maps for heavy rainfall events in recent years.
Complementing Existing Weather Stations:
Mumbai has a network of automatic weather stations that provide rainfall data at 15-minute intervals.
However, these stations do not provide information about waterlogging or flooding.
Crowdsourced real-time flooding data fills this gap and complements the existing weather station data.
Twitter Data Analysis:
Twitter users in locations with heavy rainfall and low elevations are more likely to tweet about waterlogging and flooding.
Crowdsourced data on flooding aligns consistently with rainfall data obtained from the automatic weather stations.
Specific hashtags, such as #MumbaiRains, #Mumbaiflood, and #Mumbaiflooddata, help extract relevant information from tweets.
Involvement of Volunteers and NGOs:
Around 80 volunteers from an NGO were trained to work as volunteered geographic information (VGI) participants.
They provided information about flood locations and water levels, enhancing the reliability of the crowdsourced data.
Open requests were also made to Twitter users to become volunteers and contribute flood-related information.
COMMENTS