- Heatwaves in India are becoming more frequent, prolonged and uncomfortable due to rising temperatures, humidity and warm nights, says Mrutyunjay Mohapatra from IMD.
- India’s weather forecasting system has evolved from simple temperature-based warnings to impact-based and hyperlocal forecasting that factors in humidity, persistence, night temperatures and sector-specific risks, he explains.
- In this interview with Mongabay-India Mohapatra also notes that climate change is reducing predictability because weather patterns are becoming more erratic and localised.
Heatwaves across the country continue to affect the daily life and test the resilience of people, infrastructure and public systems. Simultaneously, concerns about the potential El Niño conditions are also increasing, as India Meteorological Department (IMD) predicts a below-normal monsoon rainfall this year. As climate change intensifies some weather events, making it difficult to predict, forecasting systems are also undergoing a rapid transformation.
In an interview with Mongabay-India, Mrutyunjay Mohapatra, the director general of meteorology at IMD, discusses changing heatwave patterns, the growing role of humidity and warm nights, the evolution of impact-based forecasting, and the challenges climate change poses for weather prediction.
Mongabay: In the recent years, what changes has IMD been observing in the nature of heatwaves?
Mrutyunjay Mohapatra: Our data from 1970 to 2020 suggests a rise in the maximum (daytime), minimum (nighttime) and mean temperatures. As a result, the frequency, duration and intensity of heatwaves are increasing.
With rising temperatures, humidity in the atmosphere also increases. For every 1° Celsius rise in temperature (of planet Earth), there is around a 7% increase in relative humidity (the air can hold 7% more moisture). As a result, even if the temperature on a particular day remains similar to what it was in the past, discomfort levels are higher in a changing climate. The ‘feels-like’ temperature increases due to humidity.
There is a rise in both daytime and nighttime temperatures. If daytime temperatures exceed 40°C and nights remain warm, people do not get sufficient relief or proper sleep, which can affect the body’s physiological response. Persistence also matters. If such conditions occur for a single day, the impact may be limited, but if they persist for consecutive days, the effects become much more severe.
The impact is particularly serious for people without access to cooling facilities.
Mongabay: How has IMD’s heat forecasting evolved over the years?
Mohapatra: IMD has moved towards a more comprehensive heat assessment system that considers daytime and nighttime temperatures, humidity, wind conditions and persistence together. Based on this, we now issue district-level, colour-coded warnings in green, yellow, orange and red, for up to seven days in advance.
Earlier, the heatwave classification was mainly based on departures from normal temperatures or thresholds such as 45°C. But we realised that this alone was not sufficient because even temperatures below 45°C can become dangerous in some regions. We therefore introduced percentile-based thresholds using climatology data from 1991-2020. Now, temperatures crossing the 95th percentile (a temperature hotter than 95% of past observations from 1991 to 2020 for that five-day period) are treated as heatwave conditions, while those crossing the 98th percentile are classified as severe heatwaves, subject to other operational conditions.
At the same time, the dissemination system has also expanded significantly. About 20 million farmers receive weather information through WhatsApp. Warnings are also shared with the Indian Medical Association, ministries dealing with labour, health, education, drinking water, transport and power, along with state governments and disaster management agencies. We have also developed special products for sectors such as railways, national highways and the power sector so that vulnerable populations and critical services can prepare in advance.
Mongabay: Does IMD also receive feedback from the ground?
Mohapatra: While information about the weather is certainly reaching people, there are two important aspects: early warning and early action. There has to be a bridge between the two. Early warning-supported early action can significantly reduce vulnerability and improve disaster risk reduction.
Last year, during a heatwave in Delhi, representatives from a rickshaw puller association, along with others (NGOs), visited the IMD office. They told us that they needed heat-related information. That indicated that there was still a gap, and that vulnerable populations were not always receiving information directly or in time.
We immediately created a WhatsApp group with the association leaders and began sharing updates. The next day, they wrote the information on blackboards in the market, listing the expected temperatures and the precautions people should take. It is happening in Delhi and should be extended to other cities as well.
Mongabay: Has India established sufficient coordination among weather agencies, health systems and local administrations to translate forecasts into actionable heat response plans?
Mohapatra: Early warning and early action are inseparable, and there has to be a bridge between the two. That is what the National Disaster Management Authority (NDMA), along with IMD and other agencies, is trying to achieve.
The idea is to ensure coordination from central agencies to state governments, then to districts, blocks and finally villages.
We are also trying to improve last-mile communication. Forecasts and warnings are being shared with panchayat representatives, ward members and local officials through collaborations with ministries such as Panchayati Raj and Rural Development. Information is also being disseminated through networks like Krishi Sakhis and Pashu Sakhis.
Mongabay: When you spoke to Mongabay-India earlier about phase two of the National Monsoon Mission, you had highlighted state-level forecasting and sector-specific applications as important next steps. How do you assess the progress made so far?
Mohapatra: Earlier, monsoon forecasts were mainly limited to seasonal rainfall predictions for the country as a whole. Since 2021, IMD has started providing more granular forecasts, including monthly, seasonal and region-wise forecasts.
We have also strengthened extended-range forecasts issued every week for the next four weeks. Earlier, these forecasts mainly focused on rainfall and temperature trends. Now they also include information on possible extreme events such as heatwaves, cold waves and heavy rainfall spells, which helps sectors like disaster management, water, power and health prepare in advance.
Another major improvement has been impact-based forecasting. Earlier, warnings mainly relied on colour codes such as green, yellow, orange and red. Now, forecasts are linked with impact assessments. For example, if heavy rainfall is expected in a district, the system also estimates the likelihood of flooding, vulnerable low-lying areas, and possible impacts on agriculture and infrastructure.
IMD has also significantly expanded its flood forecasting capabilities. Flash flood guidance systems are now operational for around one lakh watersheds. Forecasts are also being shared with neighbouring countries such as Nepal, Bhutan, Bangladesh and Sri Lanka.
Urban flood forecasting systems have also been developed for cities such as Mumbai, Chennai and Kolkata, while similar systems are being expanded to Delhi and other cities. These systems combine radar networks, rainfall gauges and local forecasting tools to improve preparedness.
Mongabay: What are the areas where you think more work is still needed?
Mohapatra: Earlier, forecasts were mainly at the district level. Since 2020, we have moved towards block-level forecasting and are now expanding to the panchayat level through initiatives like Panchayat Mausam Seva.
Rainfall monitoring has also improved substantially. Earlier, district-wise rainfall monitoring relied on around 3,000 stations. Last year, we introduced a block-level rainfall monitoring system covering around 7,200 blocks. The number of weather stations is also increasing rapidly and is expected to cross 10,000 this year, further improving local-level forecasting.
At the same time, there are still important gaps. Forecast accuracy needs more improvement, especially at finer spatial scales such as blocks and villages. While overall forecast accuracy has improved by around 40-50% over the last decade, the existing forecast accuracy is around 84%. It needs to improve further by another 10-12%.
The larger vision is to move towards “Har Ghar Mausam”, where every household can access reliable weather information for the past, present and future. At present, forecasts are available at the panchayat level through mobile apps and websites, but there is a need for village-level weather infrastructure.
Another challenge is improving temporal resolution. Currently, many forecasts are issued hourly, but sectors like power require updates every 15 minutes. Climate change is also increasing the need for faster and more localised forecasting.
For this, IMD believes artificial intelligence (AI) and data-driven models will become increasingly important. Traditional physical models can take several hours to process forecasts, whereas AI-based systems can generate faster solutions within minutes, which is crucial during rapidly evolving extreme weather events.
Mongabay: How does climate change impact weather forecasting?
Mohapatra: It (climate change) is reducing predictability because weather patterns are becoming more erratic and localised. During the monsoon, for example, the frequency of heavy rainfall events is increasing, while the frequency of light and moderate rainfall days is decreasing. Severe weather is also becoming more hyperlocal.
To address this, forecasting systems need denser and more frequent observations. That is why IMD is continuously expanding its network of automatic weather stations, rain gauges and radars. At present, India has around 50 radars, and the target is to increase this to 126 in the next few years.
Climate change is also increasing the need for high-resolution forecasting and faster computing systems. IMD has already expanded its high-performance computing capacity, but forecasting at the village or farm plot level remains challenging.
This is where AI and machine learning will become increasingly important. Traditional physical models have limitations, especially for hyperlocal forecasts and sector-specific applications.
Mongabay: Around six years ago, you spoke about the possibility of a national meteorological policy. How has that discussion progressed over the years, and what do you think have been the key hurdles in moving towards such a framework?
Mohapatra: As the country grows and socio-economic conditions improve, the demand for reliable weather and climate information is also increasing. At the same time, weather information is now available from multiple sources on open platforms, which can sometimes create confusion. Therefore, people need accurate, actionable information at the right time so they can protect lives and property and plan their economic activities better.
In this context, there is certainly a need for a national meteorological policy. The World Meteorological Organization (WMO) has also been encouraging countries to develop such frameworks. Discussions on this have been progressing in India as well, and the Ministry of Earth Sciences is working in this direction. An expert group has already been formed to take the process forward.
Mongabay: There is growing discussion globally around the possible emergence of El Niño conditions this year. How is IMD assessing the current situation, and what could it mean for India’s monsoon, agriculture and heat conditions?
Mohapatra: Historically, in around 60-70% of El Niño years, India’s monsoon rainfall has been below normal. However, the country’s resilience and preparedness have improved significantly over time, both at the individual and institutional levels.
IMD has forecast monsoon rainfall at around 92% of the long-period average this year, which falls in the below-normal category. While such conditions can affect agriculture and heat patterns, there is no need for panic. At the same time, preparedness and adaptive measures remain important.
Read more: Local land-atmosphere processes influence heatwaves
Banner image: Mrutyunjay Mohapatra, Director General of Meteorology at IMD (left), and farmers at work in the heat (right). Images by Kundan Pandey and Dibakar Roy via Pexels.
Credits
