Skywatchweather.in
Skywatchweather.in
Onset is not declared because it rained. Three independent conditions must be satisfied together, and understanding them changes how the season is read.
Monalisa DuttaEvery year, towards the end of May, the same conversation begins. Heavy rain falls somewhere in Kerala, the photographs circulate, and it is widely announced that the monsoon has arrived. Sometimes that is correct. Frequently it is premature, and occasionally the rain in question has no connection with the monsoon at all.
The India Meteorological Department does not declare onset on the basis of rainfall alone. Onset is declared when three independent conditions are satisfied together, across different levels of the atmosphere, over a defined geographical area. Understanding those criteria alters how one reads the entire season, and explains why a thoroughly wet Kerala in the third week of May may still, formally, not be in monsoon.
Pre-monsoon thunderstorm rainfall over Kerala can be substantial. So can rainfall from an unrelated low-pressure system over the Arabian Sea. Before the criteria were formalised, onset declarations depended heavily upon subjective assessment, which made year-to-year comparison difficult and occasionally produced declarations which had to be revisited.
The consequences are not merely academic. Sowing decisions for the kharif crop, reservoir operation, power demand planning and a considerable volume of commodity trading are all influenced by the monsoon calendar. A definition capable of being applied identically each year was genuinely needed. The IMD adopted an objective set of criteria in 2006, and they have been applied consistently since.
The IMD monitors fourteen designated stations across Kerala, Lakshadweep and coastal Karnataka: Minicoy, Amini, Thiruvananthapuram, Punalur, Kollam, Alappuzha, Kottayam, Kochi, Thrissur, Kozhikode, Thalassery, Kannur, Kasaragod and Mangalore.
The condition is as follows. After 10 May, if 60 per cent or more of these fourteen stations record rainfall of 2.5 mm or more on two consecutive days, the first criterion is satisfied, and onset may be declared on the second of those two days.
The construction of this criterion deserves attention. It requires not one station but a majority of fourteen distributed over several hundred kilometres, so that an isolated cloudburst cannot satisfy it. It requires not one day but two consecutive days, so that a single passing system cannot satisfy it. And it cannot be applied before 10 May, which excludes the pre-monsoon thunderstorm season by definition.
This is the criterion of which most people are unaware, and it is the one which distinguishes a genuine monsoon from ordinary rainfall.
The monsoon is not fundamentally a rain event. It is a planetary-scale reversal of the wind. The entire system is driven by a seasonal transition to deep, moisture-laden westerly flow from the Arabian Sea. Rainfall is a consequence. The wind is the mechanism.
The IMD therefore requires that:
The third criterion uses Outgoing Longwave Radiation, or OLR, measured by INSAT satellites.
The principle is elegant. Satellites measure infrared radiation escaping from the Earth to space. Under clear skies, radiation originates from the warm surface and a considerable quantity escapes, producing high OLR values. Beneath deep convective cloud, radiation originates from very cold cloud tops high in the atmosphere, and substantially less escapes, producing low OLR values.
Low OLR is therefore a reliable proxy for deep, organised, vertically extensive convective cloud — precisely the structure which produces monsoon rainfall.
The requirement is that OLR fall below 200 watts per square metre within the area bounded by 5–10°N and 70–75°E. Shallow rain-bearing cloud does not produce that signature; deep monsoon convection does.
Declaration requires all three criteria simultaneously: widespread and persistent rainfall at the surface, deep and strong westerlies through the lower and middle atmosphere, and cold, deep convective cloud over the south-eastern Arabian Sea.
The normal date of monsoon onset over Kerala is 1 June. In practice it has varied by approximately a week either side in most years.
Following onset, the monsoon advances northward across the country over roughly six weeks, ordinarily covering the entire country by around mid-July, with the IMD issuing progressive advance lines as it proceeds.
This is the point I should most wish readers to take from this guide, because it is consistently misrepresented in public commentary.
The onset date has very little predictive value for the seasonal rainfall total. An early onset is not an assurance of a good monsoon, and a late onset is not a forecast of drought. The statistical relationship between the two is weak.
What actually determines the character of a season is the sequence which follows: the number and duration of break periods during which the monsoon trough shifts north towards the Himalayan foothills and the plains turn dry; the frequency with which low-pressure systems form over the Bay of Bengal and track inland; the behaviour of the Madden–Julian Oscillation; and the state of El Niño–Southern Oscillation and the Indian Ocean Dipole.
A monsoon may arrive punctually on 1 June and deliver a poor season through prolonged breaks in July and August. It may arrive a week late and finish above normal.
When the onset announcement is made, I would encourage it to be understood for what it is: a well-defined milestone confirming that the machinery has engaged. It is a starting signal, not a result.
Meteorology consultant and former Squadron Leader in the Indian Air Force, with over eleven years of experience in operational aviation meteorology, weather forecasting and flight safety support. She now advises technology organisations on AI-driven weather forecasting and decision-support systems. Consultancy services →