Roads, Tunnels and Construction Are Chipping Away at Kashmir’s Mountains: Expert

JAMMU: On August 16, a small group of University of Jammu students was heading toward Ladakh to study the ground and log coordinates. It was routine fieldwork on a stretch of road where no public memory of any disaster existed.

That afternoon, a landslide came down across the route they were driving.

No one from the university was hurt. But when Avtar Singh Jasrotia, the geologist leading the project, looked at exactly where the slope had failed, he found something unsettling. His team's own model had already flagged that precise location as a vulnerable zone in a susceptibility map that was about to be published then.

"That particular location had already been identified by our study as falling within a vulnerable zone," Jasrotia said in an exclusive interview with the Kashmir Times. "This demonstrates the importance of the susceptibility map: it can help identify areas that may be vulnerable to landslides even before a major event occurs, allowing authorities to take preventive measures and plan infrastructure development more carefully."

There’s nothing remarkable about the story - a road, a landslide, no casualties. But it is close to the best evidence a scientist can hope for - a proof that a model built from three years of fieldwork, satellite imagery and geology can genuinely predict where the ground will fail, not just describe where it already has.

Two warnings in ten days

On the morning of August 26, 2026, a section of glacier gave way high in Nepal's Langtang National Park and came down the mountain as a wall of mud, ice and rock. Nepal's disaster authorities have confirmed more than a thousand deaths and listed nearly 4,000 people as missing; across the border in Tibet, the toll stood at 16 dead and 546 missing.

It is, by most measures, one of the deadliest disasters to strike the Himalayas in a generation. Geologists watching it unfold say the world keeps systematically underestimating the danger these mountains carry and keeps paying for that misjudgment in hydropower plants, roads and lives.

In a coincidence of sorts, two days after the Nepal disaster struck, Jasrotia and his team’s peer-reviewed paper in the journal Scientific Reports asked exactly the question Nepal's disaster had raised: where, precisely, is this land most likely to fail next and can be known and done before it happens?

A warning already confirmed

The paper's authors, led by Avtar Singh Jasrotia of the University of Jammu's Department of Remote Sensing and Geology, had spent three years answering that question for Jammu and Kashmir, the northwestern anchor of the Indian Himalayas. Their conclusion, mapped slope by slope across nearly 43,000 square kilometres of some of the most unstable terrain on the earth - almost 28% of the region they studied falls inside "high" or "very high" landslide susceptibility zones. Its crux is that a slope failure isn't a matter of if, but essentially of when.

Jasrotia already had unsettling proof that the map worked. A trip to Ladakh had demonstrated that.

The ground gives way

The map's count of 28% is an average across wildly different terrain where the danger clusters around specific districts and roads. The study's field surveys and satellite mapping found the highest concentrations of risk in five districts in particular - Ramban, Doda, Kishtwar, Poonch and Rajouri.

Much of this is concentrated along highway corridors including the Jammu–Srinagar National Highway (NH-44), the Batote–Srinagar Road (NH-244), the Jammu–Poonch highway (NH-144A), the Mughal Road and the Kishtwar–Keylong route into Himachal Pradesh.

Some of that will surprise no one who has driven those roads. "Everyone already knows that places such as Samroli and Khooni Nallah are landslide-prone areas, where landslides occur almost every year," Jasrotia said. "That, by itself, is not new."

What the study adds is the places nobody had earlier flagged, like the spot en route to Ladakh with a systematic explanation for why some ground fails and other ground holds.

That explanation, he said, comes down to geology, water and gravity acting together. "Wherever faults occur, the area is generally more vulnerable to landslides," he said.

"Fault zones are often associated with river and stream channels, where water can weaken the slopes." His team built that logic directly into the map, factoring in land within 50 metres of a fault zone was rated highly vulnerable, within 100 metres moderately vulnerable, and within 500 metres less vulnerable.

When Jasrotia's team tested each factor's predictive power in isolation, road proximity came out as the single strongest stand-alone predictor of where the 669 landslides in their inventory had actually happened. On a separate statistical measure, drainage proximity was a primary factor.

What the road damages

Landslides in the Himalayas are, at their root, natural. Steep young mountains, heavy monsoon rain and constant seismic strain guarantee that slopes will fail somewhere, sooner or later. But Jasrotia was direct about how much of what his team mapped in Jammu and Kashmir isn't simply nature running its course.

"Landslides are a natural phenomenon, particularly during periods of heavy rainfall," he said. "However, human activities can significantly intensify the risk and increase the likelihood and severity of landslides."

The clearest contributing factor, in both his account and the study's own statistics, is the road. Jammu and Kashmir's mountain highways have been widened and upgraded at a rapid clip over the past decade and disturbing a slope to build one doesn't just risk that slope. The paper names seven major routes currently under construction or expansion.

"Road widening often involves cutting trees and altering the natural slope," Jasrotia said. "Cutting and excavating rocks for road construction can also destabilise slopes, particularly where the underlying formations like metamorphic rocks, the Shivalik formations or sandstone are weak or fractured. The resulting debris can accumulate on slopes and eventually trigger landslides."

Extending this argument, he added, same pressures of "the increasing population, widening of roads and industrialisation" are reshaping the region's weather too, feeding the very rainfall extremes that trigger the slopes to fail in the first place.

Jammu and Kashmir doesn't have to look far for proof, Jasrotia said, reminding of last year’s Chashoti cloud burst.

On August 14, 2025, a cloudburst tore through Chashoti, the last motorable village before pilgrims walk the final leg to the Machail Mata temple in Kishtwar district. Two flooding streams turned a crowded pilgrimage staging camp into a disaster zone in minutes. By the time search operations wound down, at least 68 people were confirmed dead, more than 300 were injured, and dozens were never found.

Experts called it a "grim warning" for the entire Himalayan belt, pointing to deforestation, road cutting and unplanned construction as forces stripping the land of its natural capacity to absorb rain.

Jasrotia cited the same disaster as a textbook case of drainage-driven risk hiding in plain sight.

"A similar cloudburst occurred about a week earlier in the same area, within one or two kilometres along the same stream," he said. "There was no habitation there, so the vulnerability of that location was largely unknown." It was only once tents, shops and thousands of pilgrims moved into that stream's path for the yatra season that a hazard nobody was watching turned into a catastrophe.

The pattern isn't confined to Jammu and Kashmir. Asked whether the Himalaya's current approach to infrastructure needs rethinking, Jasrotia pointed across the border to Himachal Pradesh, where road-building along the Mandi–Shimla route follows the Main Boundary Thrust. "Numerous landslides occurred in the region," he said. "Geology played an important role in triggering these events. The most important and urgent need of the hour is to consider both ecological and geological safety while planning and developing infrastructure in the Himalayas."

Even large, carefully engineered projects aren't immune. The Baglihar hydropower project in the Pir Panjal region is located in what Jasrotia called a geologically weak zone, because major dams are almost always built on big rivers and big Himalayan rivers tend to follow the same fault lines that make slopes unstable to begin with.

Why a map matters

Jasrotia said that a susceptibility map doesn't need to predict the unpredictable to save lives. It only needs to ensure fewer people are building, walking and sheltering in the worst of them.

The credibility of that classification rests on the study's own field validation. Of the 669 landslides catalogued using ground surveys and the FLIM (Field Landslide Inventory Mapping) App built by India's National Remote Sensing Centre, 87% had occurred exactly where the model said they were most likely to.

None of that would have mattered in Nepal on August 26. The Langtang disaster was triggered by a glacier collapse, not monsoon-soaked slopes, a fundamentally different mechanism that no landslide susceptibility map of this kind is built to catch.

But the two events share an uncomfortable lesson. Investment, tourism and infrastructure keep expanding into Himalayan terrain whose specific dangers were never systematically mapped in the first place.

Susceptibility maps like Jasrotia's are one of the few tools that exist to close that gap before, rather than after, disaster does the mapping instead.

From the map to action

Jasrotia is careful to draw a line around his own role in building a susceptibility map.

"It is up to the government to decide how seriously it takes this information," he said. "Our responsibility is to provide the scientific information; I hope the authorities will take it seriously rather than adopt a casual approach."

He’s hopeful, it might. Jammu and Kashmir's administration has constituted a committee to prepare a formal disaster-risk map, with its first meeting held roughly two weeks before this interview. It has also formed a separate "Atlas Group," on which Jasrotia sits alongside the territory's Disaster Management, Relief, Rehabilitation and Reconstruction Department, has also begun meeting. On the day of the interview itself, he said, the territory's Home Secretary was convening a further review.

Asked directly what should happen in the roughly 28% of the territory now classed as high or very high risk, Jasrotia was clear. "There should be no construction wherever there is a high level of vulnerability," he said.

"Areas falling under high and very high susceptibility zones should be treated as restricted-development or no-construction zones to reduce the risk to life and infrastructure." More broadly, he wants the map itself built into the ordinary machinery of governance, like land-use planning, building permissions and infrastructure approvals, rather than treated as a reference document engineers consult after the fact.

"That is primarily the engineering aspect," he said, when asked which specific construction techniques are riskiest. "Our role is to provide the geological perspective and identify vulnerable areas. Engineers can then suggest appropriate engineering measures, such as different types of walls, bolting, chaining and other slope-stabilisation techniques, depending on the site conditions."

A map has limitations

For all its predictive power, Jasrotia is candid about what the map cannot do, starting with precision. At its current resolution, the study offers a regional picture, not a street-by-street verdict. "To make the assessment more precise, at this point we need to carry out mapping at a 1:10,000 scale," he said.

"This is a costly exercise, but detailed mapping at this scale would provide much more accurate information."

The bigger limitation is time. A susceptibility map is like a photograph of a landscape's vulnerability at the moment it was taken. “It is not static,” Jasrotia reminds.

He elaborates. His study considers a decade of rainfall data for Jammu and Kashmir between 2014 and 2024 with a declining average. But he adds, “There is a transport of moisture from the Indian Ocean," Jasrotia explained. "When this moisture reaches the Himalayan region, it can lead to cloud formation and, in extreme cases, cloudbursts. Such intense rainfall can trigger landslides, particularly in areas with loose soil, mud and debris."

A region can be getting less rain on average and still be getting hit harder with violent bursts that overwhelm slopes mapped for a gentler climate. "Landslide susceptibility maps should not be treated as static," he said. "Future assessments need to take changing rainfall patterns and extreme weather events into account."

Jasrotia warns of another limitation. "A susceptibility map tells us which areas are more likely to experience landslides," Jasrotia said. "It cannot precisely predict when a particular landslide will occur."

Some of the region's most dangerous ground barely shows up on any map at all until it starts moving. In Doda, ground subsidence in an area of metamorphic rock forced authorities to find an entirely alternative route, and similar unpredictable instability has struck Mendhar in Poonch and the Sangaldan area of Ramban.

"There are also situations where it is very difficult to predict exactly when and where ground movement will occur," he said.

Building the warning the map can't give

If a susceptibility map answers where the vulnerability lies, the tool Jasrotia's team wants to build next is something that can map an estimated timing for the unfolding disaster.

An early warning system, as he describes it, would mean permanently installing monitoring sensors and satellite-linked instruments on the region's most vulnerable slopes, watching continuously for the abnormal ground movement that precedes a collapse. "If any abnormal movement or other warning signs are detected, the system could generate an alarm before a major slope failure occurs," he said.

"This would allow the authorities to evacuate people, restrict movement, or temporarily close roads in highly vulnerable zones. In my view, an effective early warning system is one of the most important solutions for reducing the loss of life and property from landslides."

It doesn't exist yet, and Jasrotia explains why. "An early warning system is not possible without connectivity to the relevant satellite data and a system that can process and analyse that information in real time," he said.  That would require clearances from the central government.

Until that happens, he said, "nobody can be 100 percent certain about where and when a landslide will occur."

For now, Jasrotia's recommendation to the Jammu and Kashmir administration is simple and basic: use what already exists.

"We suggest that the authorities take the information, and susceptibility maps we have provided seriously," he said. "Before undertaking any construction or infrastructure development, these susceptibility maps should be properly considered and incorporated into the planning process."

More from the interview: