A view of solar panels and wind turbines installed at the Adani Green Renewable Energy Plant in Khavda, Gujarat, October 2024.
| Photo Credit: AFP

As the sweltering heat of summer reaches its peak, our thoughts turn towards the approaching rainy season. Data collected by weather stations and rain gauges for over a century tell us that this season begins with the arrival of the southwest monsoon over Kerala on June 1, give or take a week. Weather forecasting has become more accurate in recent years, and it is predicted that the monsoon will reach Kerala around May 27 this year.

Southwesterly winds traveling over the Indian Ocean, along with strong air currents traveling over the Arabian Sea from East Africa (the Somali Jet Stream), carry moisture to our lands, refreshing our senses and lightening our moods.

In today’s world, these winds also carry with them the promise of renewable energy. An awareness of climate change has brought clarity to the urgent need for lowering our dependence on energy derived from fossil fuels. India’s position here is particularly acute. Nearly 75% of our electricity comes from coal. As a part of an ambitious plan to shift to low carbon energy, the Central Electricity Authority aims to have 121 GW of installed wind capacity by 2032, building upon an existing capacity of 45 GW.

Fossil fuel-fired power plants can generate electricity when it is needed. Renewable sources such as wind have variations, and this leads to lower capacity utilisation. Therefore predicting when wind will be in motion is critical to making the best use of investments in wind energy. The goal is to generate the maximum amount of electricity for a grid while burning the least amount of fossil fuels. Seasonal climate forecasts are necessary for this planning and start at the regional level. For example, the state of Rajasthan has very poor winds from October to December.

Monsoons are strong drivers of climate. The cool gusty monsoon winds can be predicted and modelled, just as rains can be. Cities need more power during summer, when agricultural demand is low. Power generated during the monsoons is a boon to the agriculture sector, as kharif crops  (planted in June, harvested in October) take up more electricity than the winter rabi crop. At windy locations such as the Western Ghats, a wind turbine generates 70% of its annual power output between June and September.

However, there is a great deal of variability in the velocity of surface winds during this season. Anticipating this variation is of great use in minimising deficit or surplus in power generation. This has led to the refinement of numerical weather prediction models, which work at a resolution of a few hundred metres, a kilometre, and so on. Using such models, the National Institute of Wind Energy in Chennai has developed a Wind Atlas of India, a very useful tool for planning future wind farms.

What about AI? The quantity (and quality) of high-density data from radar and satellite images has grown rapidly. Densification techniques such as Google’s MetNet3 are used for integrating this with measurements including wind speed, temperature, etc., from a relatively small number of weather stations. This allows the model to estimate wind speeds in the areas between stations, resulting in a high-resolution wind speed map derived from a small amount of directly measured data.

This article is written in collaboration with Sushil Chandani, who works in molecular modelling.