Lessons learnt from a rockfall time series analysis: data collection, statistical analysis, and applications

Historical rockfall catalogues are important data sources for the investigation of the temporal occurrence of rockfalls, which is crucial information for rockfall hazard and risk assessments. However, such catalogues are rare and often incomplete. Here, we selected and analysed seven catalogues of historical rockfalls in Austria, Italy, and the USA to highlight existing relationships between data collection and mapping methods and representativeness of the resulting rockfall records. Heuristic and simple statistically based frequency analysis methods are applied to describe and compare the different historical rockfall catalogues. Our results show that the mapping strategy may affect the frequency of the assessed rockfall occurrence and the completeness and representativeness of the related time series of historical rockfalls. We conclude by presenting the advantages and limitations of the application of different frequency-based methods for analysing rockfall catalogues and providing recommendations for rockfall mapping. We furthermore present non-parametric statistical methods for dealing with typically small rockfall datasets, which are particularly suited for the characterization of basic rockfall catalogues. Such recommendations should help in the definition of standards for collecting and using temporal rockfall data in hazard and risk assessments.