On the morning of August 3, 2026, a hillside above the Karnaprayag-Nainisain motor road in Uttarakhand’s Chamoli district simply gave way. This wasn’t some one off accident. It was part of a worrying string of geological failures that have been hitting Chamoli harder and harder. In a matter of minutes, thousands of tonnes of rock and mud came crashing down. Officials later didn’t even call the road ‘damaged.’ The videos tell a clearer story: massive amounts of rock, mud and debris racing down the steep slope and wiping out that vital stretch of road completely.
Luckily, no vehicles were crossing at the time. That’s the only reason the day is remembered as a major disruption instead of a tragedy. For the villages further along in the area locals call Kaperi Patti, though, the road that connected them to the market town of Karnaprayag, and to the clinics and schools along the way, had simply ceased to exist.
This stretch had already shown trouble before. Thirteen months earlier, in July 2025, a retaining wall near the same government technical institute had collapsed and shut the road for three days. People used an alternative route then. Over the following year, debris kept closing it off again and again, until the August 3 slide finished the job for good. Not far away, on the Badrinath highway, a landslide zone known as Ummatta had been spilling onto the road so regularly that it had frustrated engineers for more than a year. Data from the National Highways and Infrastructure Development Corporation Limited pointed to a clear trend, landslide zones across the region were expanding. The Simli-Gwaldam Highway and the Dewal-Tharali roads were also seeing long closures that left large parts of the district cut off.
Nine days after the big collapse, residents were already out protesting a plan to shut the road for four full months so it could be rebuilt properly. By August 20, according to local reports from Chamoli, the widening work had finally begun.
Nobody died on the Karnaprayag-Nainisain road in August 2026. That fact alone is worth sitting with for a moment. It’s easy to spot a bridge that gives way with people on it. It’s much harder to notice a hillside that keeps slipping, month after month, until one day it finally goes, and, by sheer luck, no one happens to be crossing when it does. For over a year, that slope had been sending out distress signals along a route mountain families depend on to reach the market and the clinic. The state’s response came only after the road had already vanished. And that raises a question we’ve heard at almost every disaster site in India: was this sudden, or was it simply the year the warnings finally ran out?
Disaster researchers draw a clear line that is easy to explain but hard to remember when you’re watching a flood unfold on the news. A hazard is simply the natural event itself, a cloudburst, an unstable hillside, a rising sea. Exposure is whatever sits in its path, a road, a village, a hospital. Vulnerability is how quickly that exposure turns into real harm, think of a poorly drained retaining wall versus a well engineered one, or a family with enough savings to rebuild versus one that has nothing left. Multiply those three factors together and you get disaster risk.
None of this means rainfall patterns aren’t changing. They are. But a hazard only becomes a disaster when human decisions place something fragile right in front of it.
A heavy downpour in the mountains is nature’s work. A road carved into an unstable slope without any drainage channel is a mistake made by humans. That is a design choice made years earlier by an agency that decided drainage could wait. A settlement spreading into a river’s active floodplain is not a hazard either. It is a permission granted, or an encroachment quietly tolerated, long before the water ever arrived.
This is why the real argument running through places like Karnaprayag, Joshimath and the Odisha coast is not really about the weather. It is about where India allows people and roads to sit in the first place. Prevention starts as a spatial and institutional decision, long before it becomes an engineering problem.
A brand new road comes with a foundation stone, an inauguration ribbon, a plaque, and a photograph in the morning papers. Maintenance offers none of that. There’s no ceremony for a retaining wall that quietly held, no ribbon cutting for a drainage channel that was cleared before the monsoon instead of after it. In maintenance, success looks like nothing happening at all, and that’s a hard thing for any political system to celebrate or campaign on.
This isn’t mainly a story about corruption, though corruption does play a role in some cases. The deeper problem is institutional. It shows up even when a project is perfectly legal and fully audited. Engineers will tell you that every asset has a life cycle: design, construction, inspection, maintenance, repair, adaptation, and eventual replacement. In much of India’s public works culture, that cycle effectively stops at construction. The finished road or bridge is treated as the end of the story, when it’s really closer to the starting gun for the long phase that decides whether the thing becomes a danger twenty years later.
You see the pattern even when no one has died. After a colonial era suspension bridge in Morbi, Gujarat, collapsed and left 135 people dead, the Uttarakhand government ordered a safety audit of every bridge in the state. The audit found 36 bridges unfit for traffic, 16 of them in Pauri district alone, another eight in Tehri. No one had died on any of those bridges. The check only happened because people had died on a bridge somewhere else. So it is not as if precautionary measures are absent. The problem is they are not mainstream and mandatory in India’s political culture.
Nowhere is this easier to follow through the years than in Joshimath, a Himalayan town about a hundred kilometres north of Karnaprayag. In January 2023, cracks started appearing on the floors, ceilings and walls of hundreds of houses across all nine wards. Digambar Singh Bisht, a resident, told Al Jazeera he woke up at 3:40 a.m. on January 3 to a loud rumble. He said it felt like the ground beneath them was about to split open. His family of five rushed outside and spent the night on the road. By morning their house was too damaged to live in. Roads and a stretch of the Badrinath highway sank along with the houses, and at least 240 families had to be moved out temporarily.
What makes Joshimath the clearest case isn’t just the cracking itself, it’s the long paper trail that came with it. A government-appointed panel, the 1976 Mishra Committee, had already warned that Joshimath sat on a deposit of sand and stones resting on an ancient landslide, and that it was unsuitable for a township. That warning sat ignored for 47 years. In the meantime the town’s built-up area nearly doubled over the following seventeen years. Geologists still argue about how much the 2023 sinking was sped up by tunnelling for a nearby hydropower project versus the instability that was already there. The operator has denied any connection, saying the tunnel doesn’t pass under the town, and the government’s own post-disaster needs assessment never listed the hydro project as a cause.
What no one disputes is that the ground kept moving afterwards. A satellite-based study found that between December 2022 and December 2024, parts of Joshimath sank by more than 30 centimetres. Compensation crawled along, only 48 families had received a total of ₹11.69 crore months into the crisis. In April 2024 India’s National Green Tribunal pulled up the state government for failing to submit any serious action plan. A ₹1,658.17 crore central rehabilitation package was finally approved in November 2023.
Further north, in Uttarkashi, a flash flood tore through Dharali village on 5 August 2025. At least five people were killed and more than a hundred others went missing after floodwaters loaded with heavy debris swept down the valley. The exact trigger is still being debated: a cloudburst, a glacial lake outburst flood, or a glacier collapse, though the weather authorities had already put out warnings of heavy rain in the area.
Himachal Pradesh tells a similar story, only on a much larger scale. The 2023 monsoon killed 428 people in the state, forced roughly 50,000 from their homes, and left behind rainfall damage that closed 128 roads at a cost of about ₹2,941 crore. Nineteen bridges were washed away and another 97 were damaged. Two years later, Kullu district’s public works department, which had taken roughly ₹275 crore in monsoon losses, still had many of its rural roads unrepaired. Officials pointed to budget shortages as the main reason. Himachal’s own 2025 human development report put the cumulative disaster losses over four years at around ₹46,000 crore and recorded about 1,700 deaths across five monsoon seasons. The state says it asked for ₹9,000 crore in central disaster assistance on the basis of a 2023 needs assessment, but received only about ₹2,000 crore.
None of this is an argument for stopping Himalayan roads, hydropower or tourism. The point is narrower. Build on these slopes with a clear sense of what they can actually bear, and treat the money for maintenance with the same seriousness that goes into the ribbon-cutting.
If the Himalayas show what happens when infrastructure meets unstable ground, the Odisha coast offers a slower version of the same failure to plan for the long term. Satabhaya, a cluster of seven villages in Kendrapara district, has been losing land to the Bay of Bengal since at least the 1970s. Its area shrank from about 350 square kilometres in 1930 to roughly 140 square kilometres by 2015. Longtime residents talk about losing fifty to a hundred feet of coastline with every cyclone. The National Centre for Coastal Research found that 33 percent of Indian coastline showed some form of erosion between 1990 and 2018.
Odisha picked Bagapatia, further inland, as the resettlement site for Satabhaya in 2008. Relocation had reportedly been talked about as early as 1992, but the actual move lagged far behind the erosion. The district administration shifted the first 571 families from Satabhaya to Bagapatia in 2018. By April 2024, officials said 17,049 displaced people had been resettled. This is a genuine and rare case of a government moving a community before every last house went underwater. It is also, on the evidence, incomplete. Bagapatia’s swampy soils have turned out hard to farm. Fishing families lost their direct access to the sea. Some of the resettled residents still cross back into the protected mangrove sanctuary around Satabhaya, because fishing there remains their only real way to make a living.
This is what ‘managed retreat,’ the deliberate, planned move of people away from land that can no longer be defended, looks like when it’s only half done. A seawall or a line of boulders can slow the erosion in one stretch, but no defensive structure can guarantee permanent safety once the geography itself is shifting. Pretending otherwise, and just rebuilding forever at public expense, would be its own kind of dishonesty toward the people living in harm’s way. Moving houses, though, is not the same as rebuilding a livelihood, a land title, or the daily run of a school. Odisha showed it could see a slow disaster coming. It has shown far less about what actually happens to people once they arrive somewhere that is supposed to be safer.
Every one of these failures gets paid for three times over. First to build the road, embankment or bridge. Then again for the rescue and relief after it collapses. And a third time to rebuild what could often have been kept standing for a fraction of the cost. Underneath those three bills sit costs that never show up in any press release. Wages lost while a market road stays closed, school terms interrupted, businesses damaged, families forced to move twice, and a slow erosion of public trust that is harder to put a price on than concrete.
The numbers are big enough to matter to any taxpayer. Himachal Pradesh alone has recorded roughly ₹46,000 crore in disaster losses over four years and has become the third most debt stressed state in India partly because of it. Joshimath’s rehabilitation package stands at approx ₹1,700 crore, more than two years after the crisis it is meant to address. In Kullu district alone, ₹275 crore of monsoon damage to roads was still largely unrepaired the following year simply for want of budget. None of these figures is the full ‘total cost’ of any disaster. They mix infrastructure damage, relief and reconstruction, categories that officials do not always keep cleanly apart.
Taken together, though, they show a clear pattern: the state finds money far more easily after a road has failed than before it does. Inauguration or reconstruction gets you a press conference. Maintenance gets you only the absence of a headline, and that is poor currency in any democracy.
Looking at places like Karnaprayag, Joshimath and Satabhaya, it would be easy to decide that the Indian state simply isn’t there. That conclusion is lazy, and it’s wrong. The state is almost never absent. Chief ministers show up within days, response teams get deployed quickly, and compensation does eventually get paid, however imperfectly. What is missing is not presence. It is sequence.
The state tends to arrive after a failure rather than before it. It finds money for reconstruction more readily than for maintenance. It issues guidelines more easily than it enforces them. It forms committees without keeping the technical expertise that outlasts any single committee’s term. Responsibility for one mountain road can be split between a Union ministry, a highways agency, a state public works department, a district administration and a panchayat, with a contractor somewhere in the middle and no office clearly owning the result if the road fails.
Call this disaster theatre. It is not a dig at the rescue workers, most of whom do genuinely hard and dangerous work. It is the cycle itself, the visit, the announcement, the cheque, the contract, that can end up standing in for the quieter, less visible work of prevention.
None of this is uniquely Indian, and it’s worth keeping that in mind before the whole argument hardens into a claim about national character. In August 2018 a section of the Morandi Bridge in Genoa collapsed during a storm and killed 43 people. At a trial years later the court heard that the risk of collapse had been raised internally eight years before the disaster. The former highways executive responsible for the bridge’s maintenance told the court he had done nothing about it and called it his greatest regret.
In June 2021 a condominium tower in Surfside, Florida, collapsed and killed dozens of people. An engineering report three years earlier had already warned of major structural damage that needed repair. The following month a town inspector told residents the building was in very good shape.
Rich democracies with strong regulators can ignore a known warning just as easily as a poorer state can. What separates an institution that actually learns is not the absence of failure, failure is close to universal, but what happens afterward. Genoa’s replacement bridge was built in about fifteen months once construction started and opened within two years of the collapse. Dozens of people, including the executive who admitted his own inaction, went on trial. Failure was followed by both consequence and capacity. That combination is one India’s institutions have yet to produce with any consistency.
None of this is an argument for paralysis. India does not need fewer roads or resettlement colonies. It needs the money and the discipline to look after the ones it already has. A serious reform programme would start with life-cycle budgeting, costing every major project for its expected maintenance, inspection schedule and eventual replacement before the first foundation stone is even laid. Maintenance funds would have to be ring-fenced so they cannot be raided for a newer project whenever budgets get tight, as seems to have happened with Kullu’s roads.
Every significant road, bridge and drainage channel would need a named institutional owner who is held accountable for how it actually performs, not just for whether it got built. It would also mean independent technical audits that happen before failure, not after, the kind of bridge inspection that, in Uttarakhand, only took place once bodies had already been counted in Gujarat. Contractors should carry real obligations well past the ribbon-cutting, with genuine penalties when failures are documented. Technical staff who know a particular slope or river should be protected from the usual routine of constant reshuffling. Hazard maps from the Geological Survey of India and the meteorological department should carry actual legal weight in construction permits, instead of sitting as advisories that planning authorities can quietly ignore.
Every major disaster should produce a public post disaster report that sets out what failed, what warnings already existed, who was responsible, and whether anything actually changed afterward. Because a disaster that keeps repeating under the same institutional conditions has not been survived. It has simply been rebuilt. Local governments would need the technical staff, the data and the enforcement power to act on all of this, rather than waiting for orders from a state capital.
None of this has to mean we’re stuck. The same country that keeps failing at basic maintenance has, in one area at least, clearly figured something out. Back in 1999 a super cyclone packing 155 mile/hour winds killed 9,658 people in Odisha and left behind $2.5 billion in damage. Twenty years later Cyclone Fani hit the same stretch of coast. This time the state managed to move 1.2 million people out of the way. Sixteen people still died and a lot of villages flooded, but officials said it could have been much worse. The UN’s disaster risk office later pointed to India’s zero casualty approach and the early warning system as the main reasons. Five years after that, Cyclone Dana came ashore on the same coast. The chief minister confirmed no one was killed after more than 800,000 people were shifted to over 6,000 relief centres.
That doesn’t wipe out what happened in Joshimath or the roads still sitting broken in Kullu. Getting people out of a storm’s path is one skill. Keeping a retaining wall upright for ten years when there’s no crisis to justify the budget is another. Still, it shows something the rest of the story can easily hide. Indian institutions can do the quiet, unglamorous, long haul preparation that stops catastrophe, when the incentives actually reward it. Odisha didn’t get better by luck or in one election cycle. Different governments kept treating shelters, sirens and drills as ongoing work, not a one off announcement. There’s no good reason the same approach couldn’t be applied to a simple retaining wall.
The Karnaprayag-Nainisain road is getting widened now. Machines have moved in on the exact stretch where the hillside came down. If they finish the job properly this time, the road might finally get the drainage and slope support that smaller slips over the past year had already been pointing to. That’s something, cautiously. But it’s still just one road. There are thousands of similar stretches across the Himalayan states sitting with the same risk, waiting for their own loud failure before anyone brings in the machinery.
The people in Kaperi Patti didn’t need a satellite report to know the road was in trouble. They’d watched it shut again and again until it simply vanished. What they couldn’t do, and only the state can, is put money into a retaining wall before it’s an emergency, keep inspections going so reports don’t disappear, or force a contractor to actually repair something instead of just cutting a ribbon. The real test of any government isn’t how fast the relief trucks arrive after a landslide. It’s how much quiet work gets done in the years when no one is watching and nothing has gone wrong yet. India hasn’t failed that test in every place. It just hasn’t passed it often enough. The harder question, why people still live, work and build on ground this unstable, is where the series goes next.
This article is part 1 of the series ‘India and Disaster Management’
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.
