Over the years, we have moved from the statistical model of forecasts, based on historical data, to a dynamical model. In 2021, we adopted the multi-model ensemble forecasts which also included forecasts on spatial distribution, instead of only categorising monsoon rainfall as normal, above normal or below normal. In this model, we also provide monthly forecasts during the June-September monsoon period.
In sum, we provide monsoon forecasts for the entire season. These are major breakthroughs to support planning and preparedness because information on spatial distribution is important and the accuracy is improving and getting better. Our understanding of the dynamic weather factors such as El Nino, La Nina and their variations over the years, the monsoon’s relationship with the Indian Ocean Dipole, the snow cover in the northern hemisphere, the sea conditions in the Atlantic or Pacific have improved.

On the basis of these models and constantly improving efficiency, we can say that since 2021, 80% of what we predict on monsoon is correct. There surely are variations from month-to-month and region-to-region, so there is certainly a huge scope to improve. But our accuracy has gone up.

Besides monsoon forecasts, we also need specific city-focussed forecasts and we have augmented the facilities in major cities and towns. We now have forecasts for 1,200 cities. We also have ‘Har, Har Mausam, Har Ghar Mausam’, which provides citizens weather information at pincode-level or cluster-level.

The days are not far away when each and every person will need to monitor weather information as there are increased risks to life and property. So, the plan is for every household to eventually have easy access to weather information.

Is a long-range monsoon forecast still relevant since the accuracy to forecast in a shorter turnaround time has improved?

We are a diverse country with varying crop patterns. Most of our farming is rain-fed and irrigation cover is not widespread. If we say there would be good rains in monsoon, water in reservoirs would be utilised accordingly for irrigation, power and industry. There is adequate planning for different sectoral works. Therefore, early stage forecasts still play a crucial role.

Recent studies have pointed out impact of dry and moist heat stress, which is a risk to everyone. In this context, does the IMD see a need to issue more public-friendly alerts on heat exposure?

There has been a paradigm shift in the approach towards heat waves as there are regular interactions between states and central agencies over preparation of heat action plans and dissemination of alerts.

So far, we have been using broad norms to declare heat waves. Based on terrains, such as hills, coasts and plains, there are different thresholds which form basis of declaring heat waves. For instance, in the plains, if temperature crosses 40 degrees Celsius and above normal by 4-5 degrees for a sustained period, we declare a heat wave.

But there are limitations in this approach. So now, we are also taking into account criteria such as warm nights, humidity, winds as well as the historical data of a weather station. For instance, when day temperatures are above 40 degrees Celsius and night temperatures are four-five degrees above normal, we declare it as a warm night and an alert is issued in our bulletins.

In the case of humidity, one feels more dehydrated and the real-feel of hot weather is different in a humid place compared to a drier place. The impact of 40 degrees Celsius in Delhi and say Bhubaneswar will be completely different, owing to humidity.

Therefore, we have also introduced humidity forecasts. At present, we are providing information on humidity for today and the next five days, in pictorial and text forms. Using these factors, we make a composite heat index to indicate what would be the real-feel of the hot conditions. All the information is available on our website and our social media. Besides, we also sent alerts.

What is the assessment on the extreme rainfall event in Dubai?

Usually, UAE and Oman experience western disturbances, which develop over the Mediterranean Sea and travel eastwards. While moving eastwards, sometimes they move towards the southern latitude. In this particular case, it was an active western disturbance which moved across the Middle East and is now moving across Iran, Afghanistan and now approaching Pakistan and India.

While the western disturbance was approaching and moving across UAE and Oman, there was an anticyclone and there was moisture incursion over the Arabian Sea. This helped in the formation of this kind of rainfall event. Rainfall above 20 cm in one day in an arid area like UAE creates problems of its own. This kind of phenomenon is not so frequent, it occurs when such weather systems interact with each other. The same western disturbance will also cause rains in J&K, Himachal Pradesh and the Indo-Gangetic plains from Thursday.

What is the status of urban flooding mapping?

Initially, we developed it for Chennai, through the initiative of the principal scientific advisor. Then we did it for Mumbai and Kolkata. Coastal cities were addressed first since the flooding is severe and involves multi-hazard events such as rainfall, ingress of seawater, river flooding and waterlogging.

After that we are looking at 5 cities, which the Centre for Development of Advanced Computing (C-DAC) is executing in project mode with the Ministry of Earth Sciences. That includes Varanasi, Pune, Ahmedabad, Bengaluru, Bhubaneswar and Pune. This is about augmenting urban meteorological services for focussed forecasts within the city.