A team from the climate studies department at the Indian Institute of Technology (IIT) Bombay has developed a hyperlocal weather forecasting system. This system is designed to provide residents with localised weather information, particularly during the monsoon season.
This system is unique as its keeps ability to provide accurate weather predictions for specific neighbourhood, streets, and areas across Mumbai. This will be a big improvement over the primary weather forecast received from the India Meteorological Department (IMD) observatories located in Colaba and Santacruz.
The city also receives weather data from over sixty automatic weather stations (AWS) and rainfall monitoring stations in Mumbai and nearby areas.
The first hyperlocal weather forecasting systems may be available to the public by mid or end June. The system will be accessible via a website linked to climate.iitb.ac.in. Until the website is live, predictions will be posted on the X profile, @ClimateIITB.
Experimental forecast system for Monsoon Rain in Mumbai developed by students of @ClimateIITB , it says onset on the day after tomorrow (Friday, 7th June, 2024) with a good rain of 10-25 mm across the city. Soon, we will come up with a portal here. #mumbairain @iitbombay pic.twitter.com/eKQbfAFYeb
— Subimal Ghosh (@subimal_ghosh) June 5, 2024
The model will provide hourly rainfall predictions for the next 24 hours for over 25 locations. It will also provide daily rainfall predictions for the next three days.
The model's first projection for Friday, June 7, indicates the start of the monsoon over the city. It predicts a good rain of 10–25 mm over the city, with Chincholi expected to receive the highest amount at 31.6 mm.
In addition to predicting rain, the portal will also function as a flood alert. Eight flood sensors are being installed throughout the city. The team is also planning to use an online form to allow users to submit the height of the floods and crowdsource flooding notifications from X.
The latest forecast model will be available in July on the IIT Bombay-incubated start-up website experiqs.tech. It will show forecasts for more than 25 places throughout Mumbai for a span of 48 hours.
The team that developed the forecasting model consists of ten members, including IIT-B students, academic members, and staff. The coordinator of the climate studies programme, Professor Subimal Ghosh, has been leading the team. They have been reportedly working on this project for a year and a half.
As per reports, the model uses a machine learning algorithm trained in a physics-based model. It uses historical rainfall data from the city over an extended period of time. The model will receive publicly accessible meteorological data across a large 100-square-kilometre area and generate forecasts for the city's hyperlocal areas.
The Brihanmumbai Municipal Corporation (BMC) and the Regional Meteorological Centre (RMC) in Mumbai have agreed to provide rainfall data from previous years in the city to the team. Once the team receives the data, they will train their model to produce hyperlocal forecasts that are more precise.
The model is created specifically for Mumbai's unique meteorological conditions. This project is experimental and non-profit.
Share
Tweet
