Upgrading weather forecasting essential

In recent years, the number and frequency of extreme weather events has increased. We are seeing erratic and high intensity precipitation in more places. India witnessed about eight incidents of extremely heavy rain (more than 204.5 mm) in July 2024. The erratic nature of this kind of downpour makes it quite difficult to make precise, accurate, and timely predictions, which is the need of the hour.

The accuracy and timely prediction, as well as the prompt dissemination of that information to the general public, is key to any disaster prevention and mitigation strategy. This becomes more important as the incidents of short-duration heavy rainfall are increasing with each passing year. In 2021, the total number of extremely heavy rainfall incidents between June and September stood at 323. This number of such incidents increased to 376 in 2022.

The short-duration heavy rain often leads to urban floods, flash floods, and landslides causing significant economic loss. The deaths of three students due to drowning in the basement of a coaching centre in Old Rajendra Nagar and nearly 400 deaths caused by a landslide in Wayanad have revealed the lethal nature of these rain-induced disasters. Not to forget the cloudbursts in Himachal Pradesh and Uttarakhand.

The tough task ahead is to make our cities and the country resilient to climate change. Our early warning and weather forecast systems, especially during the monsoon season, have to be very robust. More weather stations will have to be installed, and hyper-locality forecasts must be generated so that the people and authorities respond promptly to heavy rainfall, which is concentrated in a particular area. Equally important is to ensure that weather alerts are issued well in time and not periodically. While weather forecasting and reporting has improved significantly in recent years, the unpredictability of climate has increased too, complicating matters.

Structural challenge

In the current setup of the India Meteorological Department (IMD), the National Weather Forecasting Centre (NWFC) based in Delhi is responsible for weather monitoring and forecasting for the nation. Then come the Regional Weather Forecasting Centres (RWFCs), which do region-specific weather forecasting. There are six regions in total. RWFC issues district forecasts alongside the State Weather Forecasting Centres (SWFCs). These regional- and state-level centres are the ones responsible for city-level forecasts.

In India, weather alerts are issued at the district level, which averages 4,950 km² and can be up to 11,200 km2 in states like Telangana. This means rainfall can vary greatly within a district, making alerts sometimes too broad.

The problem could be resolved with area-specific alerts, but IMD lacks resources. Currently, they have 800 Automatic Weather Stations (AWS), 1500 Automatic Rain Gauges (ARG), and 37 Doppler Weather Radars (DWRs). To cover India fully, they need about 3,00,000 AWS and ARG and 70 DWRs.

Inaccurate predictions

The large size of our districts and the lack of enough weather stations sometimes result in inaccurate predictions. It is said to have led to the Wayanad tragedy. The Chief Minister of Kerala, Pinarayi Vijayan, speaking with the media said, “The IMD gave an orange alert for the region, meaning there will be rainfall between 115 mm and 204 mm. But in the 48 hours preceding the landslide, Mundakkai (epicentre) received 572 mm of rainfall. IMD issued the red alert around 6am, which was after the landslide had hit.” The cost of this inaccurate prediction was the death of around 225 people and 200 going missing.

In another incident on July 29, in Mumbai, schools were closed once IMD released a two-day red alert, only for the rain not to appear, creating confusion among the parents who thought the decision to close schools was unnecessary.

Another incident of inaccurate weather prediction on July 9 in Pune came as a feast to the memers. IMD predicted that Pune city would have an isolated, extremely heavy rainfall on July 9, but the city remained more or less dry. The organisation later clarified that the red alert was for only the ghat areas of Pune. However, these incidents dent the credibility of the organisation.

Time gap in alerts

Another factor that greatly affects the effectiveness of early warning systems is the timeliness of the warning. In the list of SOPs—Weather Forecasting and Warning Services, the IMD issues weather alerts at a specific time. For example, NWFC’s main weather bulletin is released at around 1300 hrs (1 pm), which is then updated periodically at 1630 hrs (4.30 pm), 2000 hrs (8 pm), and 0800 hrs (8am) the next day. In contrast, the RWFC and SWFC forecasts come only three times a day.

This large gap between each forecast could fall short very easily because the frequency of short-duration heavy rainfall is increasing. In such an incident, only a few hours of rain could cause severe waterlogging in a city. Exactly what happened on July 27, which led to the deaths of three civil service aspirants who lost their lives when the water from the road flooded the basement of Rau’s IAS study circle.

To make weather forecasts more accurate, we should adopt a hyperlocal approach. We can model it after the Weather Information Network and Data System (WINDS) being developed by the Ministry of Agriculture & Farmers Welfare. This involves setting up AWS at the Block/Tehsil/Taluk level and ARG in every gram panchayat. For urban areas, we could have AWS in each ward and ARG at the municipality level. Increasing the number of DWRs would also help. This system will improve weather alerts and help cities prepare better for heavy rains, avoiding incidents like those in Wayanad or Delhi.