Next Generation Rockfall Frequency Assessment
The Innossuise funded project RockFreq combines LiDAR data and machine learning to deliver objective, data-driven rockfall frequency assessments. The tool supports reliable risk analysis and safer decision-making in mountain regions.
Factsheet
- Schools involved School of Agricultural, Forest and Food Sciences
- Institute(s) Multifunctional Forest Management
- Research unit(s) Mountain Forests and Natural Hazards
- Funding organisation Innosuisse
- Duration (planned) 01.08.2026 - 28.02.2031
- Head of project Dr. Christine Moos
-
Project staff
Dr. Christine Moos
Meret Weh
Alexander Dominik Ezuma May -
Partner
Université de Lausanne
International EcorisQ Association
Kanton Uri
Kanton Bern
Kanton Nidwalden
Etat de Fribourg
Repubblica e Cantone Ticino
Canton de Vaud - Keywords rockfall, release, rock cliffs, frequency, machine learning, laserscanning, LiDAR
Situation
Rockfall is a significant natural hazard in mountainous regions, threatening people and critical infrastructure. Accurate rockfall hazard and risk information is therefore essential for protecting lives, supporting sustainable development, and guiding investments in risk reduction. While modern simulation models can accurately predict how rocks move downslope, estimating how often rockfalls occur and how large future events may be remains a major challenge. Today, these assessments often rely on limited historical records, field observations, and expert judgment. As a result, estimates can vary considerably and are not always transparent or reproducible. RockFreq1.0 addresses this challenge by combining high-resolution LiDAR measurements with advanced machine learning techniques. By analyzing detailed observations of rockfall activity from different geological and climatic settings, the project will create a more objective and scientifically robust approach to estimating rockfall frequencies and event magnitudes. The resulting tool will support engineers, geologists, public authorities, and infrastructure managers in developing more reliable hazard maps, risk assessments, and protection strategies. By improving the accuracy, consistency, and transparency of rockfall analysis, RockFreq1.0 will contribute to safer communities, more effective prevention measures, and better adaptation to future environmental change.
Course of action
RockFreq1.0 will develop a next-generation approach for estimating rockfall frequencies and event magnitudes using a combination of field observations, LiDAR technology, and machine learning. The project will monitor rockfall activity at selected rock walls across Switzerland using repeated high-resolution terrestrial laser scanning surveys. These measurements will create a unique dataset of rockfall events, rock wall characteristics, and block size distributions. The collected data will be analyzed to identify the geological, topographic, and climatic factors that influence rockfall activity. Machine learning models will then be used to establish robust relationships between these factors and observed rockfall frequencies. The goal is to improve the accuracy and consistency of rockfall scenario assessment compared with existing expert-based approaches. Building on these results, the current RockFreq prototype will be upgraded into a practical online tool for engineers, geologists, and public authorities. The tool will be tested and validated together with cantonal natural hazard specialists and engineering consultants to ensure it meets real-world needs. The final outcome will be a market-ready, user-friendly solution that enables transparent, data-driven rockfall hazard assessment and supports better decision-making for risk management and climate adaptation.
Looking ahead
The main outcome of RockFreq1.0 will be a publicly available online tool that enables engineers, geologists, natural hazard specialists, and public authorities to estimate rockfall frequencies and derive realistic rockfall scenarios in a transparent and scientifically robust way. The tool will be integrated into the existing ecorisQ platform, making it easily accessible to practitioners in Switzerland and internationally. The tool is primarily designed for engineering consultancies conducting rockfall hazard and risk assessments. It will also support cantonal and municipal authorities, infrastructure operators, and other organizations responsible for natural hazard management. The results can be directly used for hazard mapping, risk analyses, climate adaptation planning, and the design of protective measures. Ultimately, RockFreq1.0 will contribute to safer communities and more efficient investments in risk prevention. More reliable hazard assessments will help prioritize protection measures, improve the management of critical infrastructure, and strengthen resilience to increasing rockfall activity under a changing climate. Furthermore, the project will establish a valuable knowledge base and promote the adoption of data-driven approaches to natural hazard management in mountain regions worldwide.
