This year’s floods in the eastern state of Bihar affected over 8.1 million people in July. According to the Bihar Disaster Management Department (BDMD), the ongoing floods inundated 16 districts across the state and hit 1,310 village councils. For better crisis management, the Bihar government started deploying an artificial intelligence-based early warning system and a mathematical modelling system.
According to a news report, the technologies have helped various districts in the state carry out the evacuation of people and set up relief camps. The systems have also helped reduce the death toll. Bihar Water Resources Minister, Sanjay Kumar Jha, told reporters that the government has taken preventive steps to minimise damage from floods using the latest technology. It has been able to alert the district magistrates of north Bihar about heavy rains 72 hours in advance.
The use of the early warning system has led to fewer flood-related deaths as people can be better prepared. The district administrations and residents are getting adequate time to move to safer places, carrying out their maximum belongings and livestock. This year, Bihar started using the services of its Flood Management Improvement Support Centre (FMISC). The centre has a newly-established mathematical modelling centre (MMC) under the centre of excellence for water resources, research, and development.
The centre has developed a flood forecast model with a 72-hour lead time for rivers like Gandak, Bagmati-Adhwara, Kamala, Kosi, and Mahananda, which originate in Nepal. The centre also assesses the Ganga starting between Buxar and Kahalgaon. With mathematical modelling, AI, a personal locator beacon, and machine learning (ML), the centre assesses precipitation, humidity, temperature, and the last seven days’ hydrological data and three days’ forecasted hydrological data to provide a weather forecast, the report noted.
Earlier this month, the India Meteorological Department (IMD) announced its plan to use AI in weather forecasting, especially for issuing nowcasts, which can help improve 3-6 hours prediction of extreme weather events. IMD Director-General Mrutunjay Mohapatra explained that AI and ML are not as prevalent as in other fields and are relatively new in weather forecasting. Therefore, IMD has invited research groups to study how to use AI to improve weather forecasting. The Ministry of Earth Sciences is evaluating their proposals.
IMD uses different tools like radars and satellite imagery, to issue nowcasts, which provide information on extreme weather events occurring in the next 3-6 hours. IMD issues forecasts for extreme weather events like thunderstorms and dust storms. Unlike cyclones, predictions of thunderstorms, which also bring lightning, squall, and heavy rains, are more difficult as the extreme weather events develop and dissipate in a very short period.
Last month, over 160 people died due to lightning alone in Uttar Pradesh and Bihar. The IMD wants to better the nowcast predictions through AI and ML, which will help understanding past weather models and speed up decision-making. Technology to accurately predict and anticipate a crisis is not just necessary for governments but businesses as well. According to critical event management experts, Everbridge, when an external risk threatens operations or a supply chain, the more time there is to anticipate the threat, the more options are available to mitigate or even avoid the disruption.
For businesses, having an integrated picture of external threats and events overlaid with an organisation’s people, assets, and supply routes, along with other contextual information to enable a timely assessment and operational response is mandatory in today’s global economy.