Prabhakar Behera is 61, and for most of his life he called home one of seven villages strung along a 17 kilometre stretch of the Odisha coast known collectively as Satabhaya. About six years before he spoke to a reporter, the two acres of farmland his family had depended on turned infertile as the sea kept advancing.
When the state government relocated his family to a rehabilitation colony called Bagapatia, twelve kilometres inland, Behera moved his wife and younger sons but stayed behind himself. The household simply could not fit comfortably onto the ten decimal plot allotted to them, his elder son found work in a plywood factory in Kerala, with his younger son expected to follow after finishing school, while Behera kept fishing whenever he could.
Satabhaya’s most visible landmark, the centuries old Panchubarahi temple, has finally submerged completely and the last remains of the temple are gone forever. Villagers who had already shifted to Bagapatia would still travel back to pray there. Basant Sahani, a former resident, described the shrine as the last visible sign that their home had ever existed. When the temple collapsed into the sea, he said, their last hope that the water might stop advancing went with it. Shivendra Narayan Bhanjadeo, the temple’s trustee, noted that thirty years earlier the sea had stood three kilometres from the temple, with its loss, Satabhaya had, for all practical purposes, lost its geographical identity.
Not everyone left at the same pace, and some never left at all. An official in Rajnagar tehsil said residents of Satabhaya and the neighbouring village of Kanhupur were ready to move, while people in some other threatened hamlets nearby remained sceptical about relocating. The state resettled the first 571 families in 2018, years later, dozens more were still wading through the paperwork required to follow, with some awaiting formal orders and others still clearing land records. By most accounts, this was the first time an Indian state government had organised the relocation of an entire community threatened by sea erosion. And by the evidence of Behera’s split household and the pilgrimages back to a temple that no longer exists, it was also an unfinished one.
No one forced these families to live on a shrinking coastline. They stayed because, until it wasn’t, the sea was where the fish were, where their ancestors were remembered, and where the temple stood.
That question, why do people keep living somewhere the water or the mountain has already marked as dangerous, deserves a better answer than ‘they don’t know’ or ‘it’s fate.’ Land near a river, a coastline or a mountain road is often the only land a poor family can afford, precisely because everyone else has already priced in the risk and moved away, in a functioning land market, danger is supposed to show up as a lower price, and in India it usually does. Markets, schools and jobs cluster where people already are, and a family that leaves loses those along with the danger.
Informal settlements grow because formal, safe housing is scarce, expensive and slow to build, and because a local politician who tolerates an encroachment today gains a grateful voter tomorrow, while the official who orders it cleared gains only enemies. Land titles in these areas are often informal or disputed, which makes leaving even harder. A family with no clean title to what it already occupies has little to offer as collateral for something safer. None of this is unique to India, and none of it makes the exposure any safer. It does mean that the people living in harm’s way are rarely the ones with the power to change it.
In July 2023, the Yamuna river hit an all-time high of 208.66 metres at the Old Railway Bridge, with floodwaters lapping at the Red Fort in Delhi, the crisis pushed more than 25,000 people into relief camps. Asha Devi, 50, from Yamuna Bazaar, who had already lost nearly everything in 2023, relived the experience when the river crossed its danger mark again in a later season and she had to pack up for shelter once more.
None of this land was truly empty to begin with. A Delhi Development Authority survey found that of 9,700 hectares in the designated Yamuna floodplain zone, more than 7,300, roughly three quarters, had been encroached upon, prompting the Delhi High Court to order the DDA to clear it and fence it against future construction. Every home, market stall or mound of construction debris placed inside that space does two things at once, it puts people in the path of water, and it narrows the channel the water needs, pushing the flood a little higher for the neighbourhood next door.
This doesn’t mean every riverside structure is illegal or that every floodplain must be emptied at once. An established colony, a farm plot and a stretch of active floodway are different problems requiring different answers.
The Netherlands, which has large areas below sea level, learned a version of this the hard way after floods in 1993 and 1995 forced the evacuation of more than 200,000 people without a single dike actually failing. Its response, a programme called Room for the River, spent roughly €2.3 billion widening and relocating settlement out of dozens of stretches of floodplain along the Rhine, Meuse, Waal and IJssel rivers, giving the water back the space engineers had spent a century taking from it. India has the technical knowledge to draw the same map. It has been far weaker at turning that map into a rule construction has to obey.
Nowhere is that distinction sharper than in the Himalayan hill towns, where the shortage isn’t water but flat ground. Tourism, pilgrimage, hydropower work and the demand for second homes and hotel rooms have all pushed construction onto slopes that were never meant to hold it, and local economies have grown dependent on exactly the building activity that puts them at risk. Himachal Pradesh alone has identified more than 17,000 landslide prone sites, several hundred of them near critical infrastructure or existing habitations, according to state data.
Shimla shows what accumulates when that pressure goes unmanaged for decades. The city was designed by colonial planners for a fraction of the population it now holds, and its hillsides have absorbed the difference. A committee appointed by the National Green Tribunal concluded in 2017 that Shimla had exceeded its carrying capacity for disaster risk management altogether, and that most of its buildings stood on slopes steeper than 70 degrees, requiring the kind of contour-cutting that leaves land prone to landslides. The report did not stop construction.
In August 2023, twelve people died when a temple crowded with devotees collapsed in a landslide at Summer Hill, five more died at Fagli and two at Krishna Nagar, where an eyewitness told a reporter that construction of a slaughterhouse nearby had triggered slides years earlier, and that even after retaining walls were built, seepage continued and the problem persisted. A state science and technology official attributed that season’s landslides to hills overburdened by construction combined with soil saturated by unbroken rain.
Mountain towns need hotels, homes and roads. Pilgrims, tourists and residents all have reasons to be there. What Shimla’s slopes did not need was six more years of building on a warning that had already been delivered in writing.
Some dangers in the Himalayas can’t be pinned on a faulty permit or a missed inspection. On the morning of 26 August 2026, a chunk of glacier broke off at about 5,200 metres on the Nepal-China border, dropped more than a kilometre down the mountainside, and crashed into a river in a remote Himalayan stretch. The impact briefly dammed the water, when the debris gave way, a flood surged nearly 100 kilometres downstream through Nepal’s Nuwakot district, burying the Rasuwagadhi border crossing and the highway that links Nepal to Tibet. Within two days, Nepali police had confirmed several hundred deaths and more than a thousand missing. By month end, combined figures for Nepal and China neared 800 dead and over 3,000 unaccounted for, with the toll still being revised as this went to press. No retaining wall, building code or hazard map would have changed what happened at 5,200 metres. Nobody can regulate a glacier.
But the valley wasn’t new to this danger, and neither was the question of how to prepare for it. In July 2025, a glacial lake burst in almost the same area and swept away the bridge at the same border crossing, killing at least nine people. Alarmed, Nepal raised the risk with China the following May, at a bilateral meeting in Kathmandu where officials from both countries, including Tibet’s regional government, agreed to build a joint glacial-lake inventory and hazard maps. Afterward, Nepali officials said they never received as much data from China as they had hoped for, China’s embassy said monitoring such remote terrain was genuinely difficult for everyone, and that it had shared information in a timely manner.
Researchers who study the region note that the specific slope which collapsed in August 2026 was not being actively monitored by anyone, so better data-sharing might not have saved this particular valley. What it might do is help save the next one, which is really the only question worth asking after a disaster like this.
This is where the settlement argument stops feeling comfortable. You can drain a hillside. You can widen a road. You can rebuild a retaining wall. But you can’t negotiate with a kilometre of falling ice. In terrain like this, the only lever left is exposure. How many people, homes, bridges and highways sit in a glacier’s path when it finally moves, how quickly anyone can be warned, and whether permanent settlements get rebuilt, again and again, in the same known corridor.
India shares the same mountains and the same kind of glacial lakes, in Sikkim and Uttarakhand alike. What happened in Nepal this week isn’t a foreign curiosity for Indian readers. It’s a preview of a decision India will also have to make.
Honestly, asking people to leave a place they know, and that knows them, is one of the toughest things a democracy can do. It means facing the hard truth that no stretch of coast can be held forever. Stone walls and embankments might hold the line for a while in one spot, but once the sea or the hillside has made up its mind, nothing guarantees permanence. Pretending otherwise only puts off the reckoning.
That’s why Odisha’s decision to move the families of Satabhaya before the sea swallowed every house deserves real credit. Most places wait until the damage is already done. Still, shifting someone’s address isn’t the same as shifting their livelihood. A fisherman sent inland loses the sea itself. A farmer given a small, swampy plot loses the scale that once made farming possible. Land papers, school seats, ration cards, and clinic access don’t automatically follow a new pin code. When any of those lag behind, what started as a housing fix turns into a fresh bureaucratic maze.
The Netherlands understood this when it asked farmers in the Overdiepse polder to give up land their families had worked for generations. Officials spent years working out fair buyouts, land swaps, and compensation before anyone packed up, precisely because they treated people’s consent and their ability to earn a living as part of the engineering problem, not an afterthought. A resettlement colony that only changes the address, without securing the income, hasn’t solved the risk. It’s just moved it from the coastline into the family’s monthly budget.
Governments find it much easier to roll out relief after a disaster than to turn down a request before one ever happens. The land people most want to live on is usually the scarcest and most valuable, so saying no to a building permit, an extra floor, or a riverside stall carries an immediate personal cost for the local politician. Saying yes, on the other hand, spreads the real price out over time and across other people. When elected representatives push back against evicting their own voters, they are usually acting rationally, not out of corruption, the system simply rewards them for it.
Look at Delhi. Courts have ordered the Yamuna floodplain cleared for years, often citing the same surveys and satellite images, yet the encroachments barely shrink between one order and the next. In Shimla, the 2017 warning about carrying capacity led to no real freeze on new construction in the years that followed. None of this is a call to bulldoze the homes of the poor without thought. Enforcement that only ever hits informal settlers while leaving well-connected commercial projects untouched is not the rule of law; it is selective law, and it deserves the same criticism as having no law at all. A state that cannot say no to anyone is not being kind. It is simply delaying the hard conversation until a landslide or cyclone forces it.
Unmanaged exposure costs far more than the headlines ever show. Delhi in 2023 shut schools and businesses for weeks, left a barrage gate stuck open, and forced families like Asha’s to pay for temporary rooms while their own homes slowly dried out. In Himachal, the landslide season regularly blocks the roads that farmers, students and daily wage workers rely on. Migrant labourers have even died sleeping in flimsy huts on shaky ground, simply because that was the only spot they could afford near their jobs.
Economists have a name for what is really going on in a floodplain settlement or an unstable hillside colony, socialised risk. A builder, a landlord or a local patron can make money putting a structure where it does not belong. When the flood or landslide finally arrives, the bill, for rescue teams, compensation, lost school days and rebuilt roads, gets spread across every taxpayer in the state.
Making that risk visible early, through insurance rates, land prices and building permits that actually match a proper hazard map, would not magically fix the housing shortage that pushes people onto dangerous land in the first place. But it would stop the state from treating the danger as if it comes free of charge.
A rule based approach does not mean shutting down development. It simply means sticking to a few clear decisions and applying them the same way every time.
Hazard maps from the Geological Survey of India and the Central Water Commission should carry real legal weight in every building permit. Right now they mostly sit as advisory layers on websites that almost nobody checks. Floodplains need to be graded by actual risk, an active floodway, an occasional flood zone, an established settlement, a farm plot, rather than treated as one big forbidden zone or one big free for all.
Mountain towns need firm ceilings on how densely they can build. Those limits should be tied to slope, drainage and how quickly people can evacuate, and they should be enforced by planners and engineers who actually work inside the local administration, not called in after the damage is done.
India also needs a standing, properly funded protocol for managed retreat. A community watching its coastline erode or its hillside start to slide should already know how the move will be paid for, where the replacement land will come from, and how livelihoods will be rebuilt, instead of having to negotiate every answer from scratch once the crisis has already begun. And whatever rule applies to a settlement on the Yamuna’s bank must apply equally to the hotel built on the same floodplain by someone with better lawyers. Otherwise it is not really a rule at all.
None of this means India is stuck repeating the same mistakes forever. Odisha’s decision to move an entire community before the sea took everything, however imperfect the process, is already being pointed to by researchers and officials as a possible model for other villages along the eastern coast that face the same advancing water. A simple decision to treat exposure as something the state can plan against years ahead, rather than something it can only mourn after the fact. India already has planners, engineers and hazard data that would be respected anywhere in the world. What it has been far less willing to do, so far, is let that expertise override a convenient local land deal or an unpopular ban on new construction.
Prabhakar Behera and his former neighbours are still out fishing along a coastline that has already taken their farmland, their temple, and most of the map they would have recognised from childhood. His wife and sons live in the resettlement colony built to fix exactly this problem, and so far it has only half-succeeded. None of those 571 families did anything reckless to end up here. They simply lived where their parents had lived, farmed what their parents had farmed, worshipped where their neighbours worshipped, and trusted a coastline that, for generations, had held steady. The real failure was never his. It sat in every land record, hazard map and building permit that kept treating the village as permanent despite the sea declaring its plans almost every year.
A disaster turns political long before it becomes visible, in a map that gets filed away, a permit that gets stamped, a warning that never gets enforced, a safer plot that is promised but somehow never quite arrives. India will need more than sympathy to change that sequence. It will need the same rules applied the same way, to the poor and the powerful alike, before the next stretch of coast or hillside makes the choice for everyone. Whether the country can now turn this hard domestic reckoning into a credible voice on resilience abroad, through the very international coalition it helped create, is where this series turns next.
Lower per-tonne freight costs, sharply lower fuel use, multimodal terminals, Jalvahak incentives and private investment are shaping a system in which heavy cargo can move by river for the longest, most efficient leg of journeys.
Ask who owns a stretch of national highway, or an embankment, or the tunnel wall at Silkyara. The honest answer is often several departments, none of whom will admit it first. A road, a bridge, a drainage channel should each belong to one named office, the kind that gets credit when it holds and blame when it doesn’t.
India’s goal now should not be just to become a ‘Builder’s Navy’, but a ‘Fast Builder’s Navy’. This means freezing designs, placing large orders together and never letting shipyard assembly lines sit empty.
Selected issues, clear explanations and thoughtful perspectives—delivered when we publish.
