A pre-monsoon storm that struck Delhi on Tuesday evening and continued through the night reached 128 km/hour at Pusa — higher than the 120 km/hr recorded at Palam airport hours earlier, which had already diverted at least two flights and delayed more than 400 others. Across the city, residents woke on Wednesday to fallen trees and blocked roads, with localities including Hauz Khas, Defence Colony, Panchsheel Park and Vasant Kunj among the worst affected; photos posted to X showed uprooted trees that had crushed boundary walls and brought pavements up with their roots.
<figure class="art
Most of that destruction arrived with almost no rain.
The mechanism behind the storm, and behind the unusual frequency of such events this pre-monsoon season, points to a specific and increasingly documented set of atmospheric conditions over northwest India.
The ingredients
Two things shaped Tuesday’s atmosphere. The first was extreme heat. At 43.5°C — four degrees above normal and the highest June reading so far — the ground was warming far faster than the upper atmosphere.
This steepened what meteorologists call the lapse rate: the rate at which temperature drops with height, or the temperature gap between the scorching ground and the cold sky above. The wider that gap, the more unstable the atmosphere becomes and the more energy it stores.
A pocket of hot surface air, once it starts rising, stays warmer and lighter than the air around it at every level. It keeps going up. It accelerates. This stored upward force is measured as CAPE (Convective Available Potential Energy) — in effect, how much explosive energy the atmosphere has built up.
Research on a comparable Delhi pre-monsoon dust storm in May 2018 recorded CAPE values of 2,696 joules per kilogram, with a lifted index of -8.98, both indicating severe convective forcing (Chakravarty et al., Agricultural and Forest Meteorology, 2021). Values above 2,500 J/kg are typically associated with severe storm conditions.
Even extreme CAPE does not guarantee a storm. Another quantity — CIN, or Convective Inhibition — acts as a lid that helps keep the instability bottled up. High CIN values typically suppress cloud development, even if CAPE—the energy that fuels updrafts—is high.
Through Tuesday, surface heating would have steadily eroded the CIN lid. By late afternoon, it was gone.
Ashwary Tiwari of IndiaMetSky said CAPE essentially acts as the ‘juice’ for a storm of this intensity. “Essentially, it is an unstable environment and just needs a spark for thundercloud development. Higher the energy, the greater it helps in development of larger and stronger storms,” he said, stating under it, warm moist air rises rapidly upwards. “CAPE and CIN work against each other in that sense in the region,” he added.
The second ingredient was moisture, fed into the heated column by a cyclonic circulation — an anticlockwise spiral of winds — drifting in from the northwest, near Pakistan. Moisture matters for a reason beyond simple humidity. It lowers the Lifting Condensation Level — in other words, the altitude at which rising air cools enough to condense into clouds. A lower cloud base means a pocket of rising hot air hits condensation sooner; at that point, condensation releases heat back into the rising air, making it climb harder and faster still.
So, the surface heat provides the stored energy; the moisture lowers the trigger threshold and amplifies the updraft once it fires.
A surface-level trough — a corridor of low pressure running across Rajasthan, Haryana and Delhi — acted as a funnel, drawing moist air inward and upward into the base of the developing column. “High heat and moisture caused instability,” said Tiwari.
“A low surface-level trough also persists, which will eventually turn into the axis of the monsoon,” he said.
M Mohapatra, IMD director-general, told HT that such storms are not uncommon during the pre-monsoon, but their intensity can depend on instability in the atmosphere. “Typically, there are four factors which determine the type and intensity of such storms. The first is intense heating and we saw that for the last two days. The second is moisture. The third is unstable atmosphere and the fourth is a trigger, which is typically a weather system. In this case, it was the cyclonic circulation. If there is adequate moisture, we get ample rain, but typically, we see drier storms across northwest India. There is a higher chance of rain late in the night, when temperatures dip,” he said.
Also read:How good or bad are the monsoon rains so far? | Number Theory
The detonation
The atmosphere does not build towards a storm gradually. It holds, then detonates.
By 5:30 pm Tuesday, the lid had broken. A cumulonimbus — the deep, anvil-topped storm cloud, capable of reaching 15 km into the atmosphere, higher than most commercial aircraft fly — developed within the following hour, its updraft pulling warm moist surface air upward at tens of metres per second. Precipitation formed high in the cloud.
The dry thunderstorm
Here is the counterintuitive core of Tuesday’s occurrence.
The air below the cloud base over northwest India in the pre-monsoon is extremely hot and dry. Precipitation falling from the cumulonimbus dropped into this layer and began evaporating before it could land — a phenomenon called virga, visible as wispy grey curtains hanging beneath a cloud base that dissolve mid-air.
As the precipitation evaporated, it stripped heat from the surrounding air column, cooling it rapidly. Cold air is dense and heavy; this mass began sinking fast. Falling rain and ice particles dragged the air downward with them, accelerating the descent further. The result was a downdraft: a column of cold, dense air driving toward the surface like a piston falling from altitude. Research on the May 2018 Delhi dust storm confirmed this downdraft-driven mechanism as the defining feature of this class of storm (Banerjee et al., Journal of Geophysical Research: Atmospheres, 2021).
At the surface, the descending column hit the ground and spread outward rapidly in all directions — like cold water poured onto a flat table. This spreading mass is the cold pool outflow, the gusts advancing all across the ground, lifting loose topsoil as it goes. The wall of dust was the visible footprint of a downdraft that had evaporated its own rain on the way down.
The rainfall data from overnight makes the mechanism legible.
Palam, which recorded the evening’s highest wind speed at 120 km/hr, received just 0.1mm of rain across the entire night. In other words, the most violent outflow arrived precisely where almost no rain landed — the storm was strongest where it was driest. Further from the outflow’s epicentre, rainfall increased: Safdarjung logged 9.6mm between 11:30 pm and 2:30 am and a peak gust of 74 km/hr; Lodi Road recorded 7.4mm; Pusa, which saw the night’s highest wind speed of 128 km/hr, received 4mm. “No significant fall in temperature was reported,” said India Meteorological Department (IMD) scientist Krishna Mishra of the evening storm, “due to it mostly being a dry thunderstorm.”
The gradient in wind readings across stations also reflects terrain. “There are limited obstructions around the airport,” noted Mahesh Palawat of Skymet, explaining Palam’s exposure. Open, flat ground gives the outflow no friction to slow it.
Also read:The long arm of El Nino that suppresses monsoon rains | Number Theory
The Andhi, and what the season is showing
The Andhi needs no introduction to most readers in Delhi — the pre-monsoon dust storm that arrives with violent, darkening suddenness is one of the city’s most recognisable seasonal events.
What the meteorological literature, going back to P.V. Joseph’s foundational study in Mausam in 1980, establishes is the mechanism: that the Andhi is driven not by surface winds but by a downdraft — the cumulonimbus cloud’s cold pool outflow described above — which is why its force and its rain so often arrive in different places. Historical records put Delhi’s average at roughly eight Andhis a year (Joseph, Mausam, 1980). Sunday’s 101 km/hr at Palam was a separate episode within the same pre-monsoon window.
The broader pattern of the 2026 pre-monsoon warrants closer attention. Western disturbances — cyclonic storm systems that travel eastward along the subtropical jet, ordinarily most active in winter — have doubled in frequency during June over the last 20 years, driven by a delayed northward migration of the jet (Hunt, Weather and Climate Dynamics, 2024), a band of westerly winds. More western disturbances in June mean more cyclonic circulations available to deliver moisture into the heated plains, and more frequent convective triggers of the kind that fired on Tuesday.
Also read:Why Delhi’s water crisis keeps returning to the same chokepoint
This intersects with a well-established finding in the heatwave science literature: that northwest India’s extreme heat episodes — the sustained 46-48°C events — are consistently linked to an upper-level anticyclone, a high-pressure system that suppresses convection, seals heat under clear skies, and blocks moisture from entering.
When that system is firmly established, temperatures build uninterrupted for days on end. When it is absent or weak, moisture penetrates, and convective storms fire at lower thresholds (Ratnam et al., Scientific Reports, 2016).
