In a few weeks the winds over the peninsula will reverse: southwest monsoon that has watered the western coast since June retreats, and moisture-laden northeasterlies run across the Bay toward our shore.
For much of India the rains are ending.
For Tamil Nadu, coastal Andhra Pradesh and Odisha, the hardest weather of the year is only beginning.
This year that reversal carries an added edge.
The 2026 southwest monsoon ran below normal under a strengthening El Niño the Pacific warming that suppresses the summer rains. But El Niño does not treat the country evenly.
Over the southern peninsula, a weak southwest monsoon is often followed by an active northeast monsoon, and with a positive Indian Ocean Dipole now developing in the Bay, the setup resembles seasons we would rather not repeat.
The 2015 deluge that drowned Chennai unfolded under precisely this signature.
None of this is destiny but when the signals lean this consistently one way, prudence is not alarmism. It is planning.
The eastern seaboard needs no reminding of what the Bay can send. In October 2025, Cyclone Montha crossed the Andhra coast near Kakinada and the response spanned three states; weeks later, Cyclone Ditwah forced red alerts across northern Tamil Nadu and Puducherry. Before them, Michaung turned Chennai’s streets into canals in 2023.
The forecasts point to a near-normal count of cyclones but the count has never been the danger. A single severe system, tracking into a delta at high tide or stalling over a city’s choked drainage, does more damage than a season of weak lows; and the models expect systems to favour the southern Bay, closer to us.
Each state carries a different vulnerability. Tamil Nadu’s signature risk is the city behind the coast Chennai’s drainage, encroached waterways and low-lying settlements, where an ordinary spell becomes a flood. Andhra Pradesh’s exposure is its long, low coast and deltas at sea level, where storm surge, not just rain, is the killer. Odisha remains the benchmark: since 1999, it has built a near-zero-casualty response on shelter networks, a standing state force in ODRAF and mass pre-emptive evacuation, moving tens of thousands before a landfall as routine.
India’s forecasting is now world-class: we know, days ahead, where a system will form and where it will strike. The failures almost never happen in the forecast they happen in the last mile, between a warning issued in a control room and a fisherman who never got it, a shelter left locked, a drain never desilted.
Closing that gap is the work of the weeks left, not the hours after a warning. Desilt the storm water drains before the first depression. Audit every shelter for water, power and access. Map the vulnerable to the household coastal hamlets, low-lying colonies, those who cannot self-evacuate. Run the table-top exercises now, so that no agency meets another for the first time in a crisis.
Push the warning out in Tamil, Telugu and Odia, through every channel people check, until it is believed. Preparedness done in October is measured in lives; preparedness attempted in December, in regret.
The National Disaster Response Force (NDRF) has changed the arithmetic of these storms, and its most important contribution is one the public rarely sees. The visible part is deployment: the force no longer waits for a disaster to declare itself but moves ahead of the storm.
For Montha, teams were pre-positioned across the vulnerable districts of Andhra, Odisha, Tamil Nadu and Puducherry days before landfall each carrying flood-water rescue and collapsed-structure capability, with airlift on standby to surge as the track shifted.
The clearest proof came only weeks ago, close to home. When heavy rain battered Kerala in early August, NDRF teams were not rushed in after the water rose 11 teams had been pre-positioned across the state under the pre-monsoon mechanism, before the first distress call.
What held it together was the discipline the eastern coast now needs: rainfall read in real time against reservoir levels, so rain pouring into brimming catchments was seen for what it was; a common operating picture shared between the weather and dam agencies, the state control room and every force in the field; and the multi-agency table-top exercises run before the monsoon, so those coordinating on the day had already rehearsed it.
The truest measure was not how many were pulled from the water, but that over 3,000 were moved to safety before they needed to be.
The larger, quieter contribution is force multiplication.
A national force will never be the first hand on a stranded family’s shoulder that is always a local. Its enduring value lies in building that local capacity: training the state forces, civil defence and Aapda Mitra volunteers, and rehearsing with them before the season so the trained responder is already in the village when the water arrives, and the NDRF is the reinforcement, not the sole rescuer.
That is the standard worth setting. Odisha has shown a well-drilled state can meet a super-cyclone and count almost no dead; with the same warnings and forces, its neighbours can hold themselves to the same.
The winds will turn in a few weeks, as they always do.
The forecasts are the easy part now, and they are already speaking. The only question that matters is whether our preparation turns with the wind or waits, as it too often has, for the water to make the argument for us.
(The writer is Assistant Commandant, National Disaster Response Force. The views expressed are personal.)