Blog - The road from Rishikesh to Joshimath, more than one landslide per kilometer

Roads increasingly pervade the Himalayas and enable access to the remotest villages. They are important lifelines for the economic development of rural areas and tourism, but they also interfere with a fragile landscape. High uplift rates, rapid river incision, trenching rains, earthquakes and weak bedrock render many regions in the Himalayas prone to landsliding. Road construction often exacerbates the hazard of landslides as hillslope undercutting and blasting further reduce slope stability.

As part of our Co-PREPARE project, we had the chance to travel to Chamoli, Uttarakhand, in Oct. 2022 to visit the area hit by the Chamoli Disaster. We travelled along NH-7, the national highway connecting Rishikesh and the holy shrines of Badrinath and Kedarnath. The road is highly frequented as many pilgrims travel to visit the temples. We followed the road until Joshimath, a town located at ~1900 m upslope to the confluence of Dhauliganga and Alaknanda. Joshimath is currently in the news as it has been built on old landslide deposits and is facing severe problems due to land subsidence (see article by Kavita Upadhyay in India Today and blog posts (1, 2) by Dave Petley).

On our way, we recognized a massive number of landslides along the NH-7. These landslides were not necessarily big, but many of them partially or fully blocked the road (see images below). Most of them had been triggered by a period of anomously high rainfall just before our travel. Hence, the debris had not yet been fully removed, yet. Being stuck in the car during the 9-hour ride to Joshimath, we thus spontaneously decided to inventory the landslides using hand-held GPS.


In our recent study (manuscript is currently in discussion in NHESS), we analyze the inventory landslides along NH-7. In short, we aim to identify the controls on the spatial locations of landslides. As opposed to most susceptibility studies, we model landslides as network-attached point pattern. If you follow this blog, you’ll know that TopoToolbox has some tools to study point patterns along stream networks (PPS), and yes, we have simply applied the tool to a road, which works equally fine.

We find that landslide occurrences can well be explained by the spatial distribution of rainfall, slope and lithology. In particular, low-grade metamorphic rocks of the Siwaliks and Lesser Himalaya are prone to shallow landslides. Yet, this is not the full story. Large stretches along the road were widened during the last 10-15 years. Below is a comparison of the NH-7 a few kilometers away from Rishikesh. Vegetation removal and undercutting of slopes have left a swath around the road extremely susceptible to mass movements which explain the overall high density of landslides.

rishikesh-to-joshimath_3 Comparison of the NH-7 shortly after Rishikesh for two images taken in 2005 and 2022.

Visual satellite image interpretation reveals that these landslides are often reactivated. 20-40% of road-blocking landslides in our study are simply reactivated and it is likely that they will dump debris on the road again during upcoming periods of extreme rainfall. To this end, our data shows that road construction – if not done properly – will have a long-lasting effect on landslides and road safety.