Italian Science Fund · FIS 3 Starting Grant · 2026–2031
Hosted at OGS, Trieste · Principal Investigator Sansar Raj Meena
NATURA studies how natural processes and human activity together shape landslide risk in mountain regions. It measures where and how often slopes fail, who and what is exposed, and what makes communities vulnerable, in the Himalaya of India and Nepal, the Western Ghats and the Northern Apennines, with further study areas in Africa and South America.
Photo: Perarolo di Cadore, Belluno Dolomites. Drone survey, February 2025.
The project in numbers
The project
Roads, settlements and land-cover change alter where slopes fail, and people in turn move into or away from landslide-prone ground. Most risk assessments treat these separately. NATURA measures them together.
Quantify how landslide hazard interacts with human-driven change: roads, settlements and deforestation.
Characterise the socio-economic factors that make people exposed and vulnerable to landslides.
Project landslide risk under climate and socio-economic scenarios, for planning and adaptation.
Research
Each strand supplies one part of the risk equation, and each is checked against the other two. Mapping is done on slope units, the natural hillslope units of the terrain.
Multi-temporal landslide inventories from a decade of satellite imagery. From them: susceptibility, how often slopes fail, how large the failures are, and how road construction changes landslide occurrence.
Led by Ankit Singh
How built-up area and roads have spread across landslide-prone slopes through time, and which socio-economic conditions explain who suffers losses.
Athif Sayyaf's doctoral research
Drone LiDAR and field campaigns, lithological and structural controls on slope failure, and satellite radar to detect slow-moving landslides. The field data validate the maps.
Led by Saurabh Singh
Study regions
The same workflow is applied in very different settings, from monsoon-driven Himalayan valleys to fast-growing cities on steep tropical slopes, to learn what transfers between them and what does not.
Indian Himalaya
Steep, tectonically active terrain with intense monsoon rainfall and fast road and settlement growth.
Italy
The May 2023 Emilia-Romagna rainfall event triggered tens of thousands of landslides on long-settled hillslopes.
North-East India
The wettest part of the Himalayan arc, with fast-growing hill towns and new road and rail corridors.
Nepal
A high-mountain valley where an ice–rock avalanche on 26 August 2026 sent a debris flow downstream.
Western Ghats, India
Steep, deeply weathered slopes where monsoon rainfall triggers landslides on cut slopes, plantations and settled hillsides.
Africa
Densely populated tropical highlands where land-use change, roads and growing towns meet landslide-prone slopes.
South America
Cities and towns built on steep slopes, where intense rainfall strikes hillside neighbourhoods.
Roadmap
The milestones overlap: validation in the field starts while the models are still being built, and the scenarios build on both.
2026 – 2029 now
Landslide inventories, hazard models, exposure and vulnerability for the study regions.
2026 – 2029
Field and drone campaigns, stakeholder interviews, and historical aerial and satellite imagery.
2029 – 2031
Risk projections under climate and socio-economic scenarios, and decision-support tools for disaster-risk agencies.
Research programme to 2028
From the field
Drone surveys show what satellites cannot: the scarp, the material, and how close the road and the houses are.
Team
NATURA is based in the Geophysics Department of OGS, at Borgo Grotta Gigante near Trieste. The doctoral research is carried out with the University of Padova.
Principal Investigator
AI and Earth observation for geohazards. Researcher at OGS.
Postdoctoral researcher
Landslide inventories, susceptibility and hazard. PhD, IIT Mandi.
PhD student, University of Padova
Exposure and socio-economic vulnerability to landslide hazard.
Researcher
Fieldwork, structural geology and InSAR. PhD, Banaras Hindu University.
Outputs
Papers are published open access, with the data needed to check their results. Inventories, maps and code are released as each study is completed.
Integrating geomorphology-based terrain segmentation with machine learning for landslide susceptibility assessment in the Darjeeling Sikkim Himalaya, India. Singh, S., Raju, A., Rosi, A., Singh, R. P., Kumar, S., Floris, M., Meena, S. R. Geosciences, 2026. doi:10.3390/geosciences16100426
Preprints, not yet peer reviewed
Mapping landslide change and volume with repeat airborne LiDAR after the May 2023 Emilia-Romagna event: slope-dependent detection limits and lowered areas. Meena, S. R. EarthArXiv preprint, 2026. doi:10.31223/X58R5C
Brief communication: NISAR pixel offsets measure but do not statistically resolve pre-failure motion at the 26 August 2026 Bhote Koshi ice–rock avalanche, Nepal. Meena, S. R. EGUsphere preprint, under review for Natural Hazards and Earth System Sciences, 2026. doi:10.5194/egusphere-2026-5462
Masked autoencoding (MAE) outperforms joint-embedding prediction (JEPA) for frozen-probe very-high-resolution landslide segmentation. Meena, S. R., Tang, X., Catani, F. EarthArXiv preprint, 2026. doi:10.31223/X56N4B
Every NATURA paper is deposited open access.
Released with a DOI when the related paper is published.
Write to smeena[at]ogs[dot]it.
Funding
NATURA, “Assessing Natural and Anthropogenic Drivers of Landslide Risk in Mountainous Regions”, is funded by the Italian Science Fund, FIS 3 call (finanziato dal Fondo italiano per la scienza – Bando FIS 3), Ministry of University and Research (MUR). Project FIS-2024-07495, CUP F53C25000810001. Host institution: OGS. Views expressed here are those of the project team.
Get in touch
We welcome collaborations on landslide inventories, field sites, census and survey data, and on social vulnerability in mountain communities. Write to smeena[at]ogs[dot]it.
