In the Bay of Bengal, the weather does not arrive as an abstraction. It arrives across salt flats and mangrove edges, across fishing grounds, rice paddies, and low-lying camps where tarpaulin is the nearest thing to a roof. By early 2023, the coast of western Myanmar and the refugee settlements just over the border in Bangladesh had already been living with a hard truth: the land was vulnerable before any storm had formed.
Rakhine State, on Myanmar’s western shoreline, is a place where the sea and the monsoon have long negotiated the terms of daily life. Sittwe, the state capital, sits low enough that storm surge is not a theoretical hazard but a recurring memory. The city’s port, breakwaters, and drainage channels existed to moderate that danger, yet the broader system was fragile. Roads outside the urban center were narrow and easily cut. Rural housing was often light-framed, built from timber, bamboo, corrugated metal, and whatever could be repaired cheaply. In the camps and informal settlements around Cox’s Bazar in Bangladesh, the density was far greater and the shelter weaker still.
That vulnerability had been compounded by displacement. More than a million Rohingya refugees lived in Bangladesh’s Cox’s Bazar district, concentrated in camps that UNHCR and aid agencies had spent years trying to stabilize against fire, monsoon floods, and cyclone winds. The shelters were engineered for emergency residence, not permanence. The ground had been stripped and terraced; drainage was improvised; access roads were narrow; slopes could fail. The camps had warning systems, but warnings are only as good as the structures people can reach in time. In practical terms, that meant every alert carried an implied question: could a family leave a bamboo shelter, navigate a muddy track, and reach a reinforced building before the weather closed the route?
For those inside the settlements, ordinary life before the cyclone was lived in a narrowed geography. Water had to be collected, food distributed, children kept within sight, tin sheets tied down, and bamboo walls patched after every hard rain. Aid agencies had pre-positioned supplies in anticipation of the storm season, and local volunteers had practice in moving families to reinforced structures. Yet the deeper blind spot was structural: the camps existed in one of the most cyclone-prone basins on earth, but the people inside them had very little latitude to relocate, and almost no durable building stock to absorb a major strike. The same terrain that made emergency relief possible—dense concentration, mapped footpaths, managed distribution points—also created points of failure when wind and water arrived together.
The same was true in Rakhine’s towns and villages, though for different reasons. The region had been shaped by years of conflict, contested administration, damaged infrastructure, and recurring restrictions on movement. Public trust in official protection was thin. Even when a warning exists, it lands in a landscape where fuel is scarce, transport uncertain, and many households cannot afford to abandon what little they have for an uncertain evacuation point. The system meant to protect people was therefore not just meteorological; it was political, economic, and logistical. Its blind spot was the assumption that a warning can function without mobility. A forecast can be accurate and still fail if roads are blocked, vehicles are absent, or a family has no safe destination to reach.
There were, nonetheless, islands of competence. Bangladesh’s disaster management apparatus had spent decades learning from past cyclone failures. The country had built a reputation for mass evacuation and community-based preparedness. In Myanmar, international agencies and local responders had for years practiced emergency coordination in a far harder environment, where access was constrained by bureaucracy and insecurity. These systems did not eliminate risk; they reduced it at the margins. Their success depended on time, trust, and the assumption that a storm would give notice. That was the quiet wager beneath the entire pre-landfall response: that the atmosphere would reveal its intentions in time for human systems to react.
But the Bay of Bengal had already made that wager look precarious. Over warm ocean water, the season had already begun to load the dice. Sea surface temperatures in the northern Bay of Bengal were high enough to feed rapid intensification, and the basin had a history of producing compact but lethal cyclones that could organize quickly and leave little margin for error. Mocha would not be the first storm to exploit that geometry. It would, however, arrive at a moment when the human geography beneath it was especially exposed. Warm water offshore, exposed slopes inland, and densely packed shelters along the edge of the sea formed a chain in which any weak link could become decisive.
The camps had become a city of shelters. The coast had become a corridor of dependence on fragile roads, shore embankments, and warnings from meteorologists. Everyone who had lived through previous storms understood the rhythm: the first advisories, the sea’s restlessness, the darkening sky, the anxiety over whether the system would bend north or west. No one yet knew that this one would deepen over exceptionally warm water and become one of the strongest cyclones ever observed in the Bay of Bengal. In that sense, the danger was not hidden at all; it was visible in the type of housing, the topography, the drainage channels, and the ordinary arithmetic of evacuation capacity. What was hidden was how quickly those visible weaknesses would be forced to prove themselves.
For disaster planners, the pre-storm period is always a forensic pause. Maps are reviewed, shelter lists checked, stockpiles counted, transport routes re-seen as lifelines rather than roads. In a place like Cox’s Bazar, where the refugee camps had long been monitored by UNHCR and other agencies, that pause mattered. The camps’ warning systems and reinforced shelters were the product of years of adaptation to exactly this sort of risk. But adaptation can only go so far when the basic conditions remain unchanged: steep slopes, light construction, congested settlement, and limited room to move. The same was true in Sittwe and the surrounding coastal districts, where the infrastructure existed, but only just.
What made the coming disaster so dangerous was not only its strength, but its target. A powerful cyclone over open water is a meteorological event. A powerful cyclone over a coast packed with refugees, fishermen, farmers, and families in light shelters is a human test. By the time the first official warnings were posted, the region’s defenses already showed the cracks that would matter most. The question was no longer whether the storm could be named, tracked, and measured. The question was whether the existing system—its roads, embankments, shelters, and channels of authority—could absorb what the atmosphere was preparing to deliver. Then the pressure began to fall, and the sea started to signal what the maps had not yet fully admitted.