LANDSLIDE INVENTORY MANAGEMENT AND USAGE IN THE NORWEGIAN EWS

The Norwegian Water Resources and Energy Directorate (NVE) was given the responsibility of initiating a national early warning system (EWS) of water related landslides (debris flows, soil slides, slush flows) in Norway back in 2010. The EWS became operational in 2013 operating nationwide at a regional scale. The development of the EWS is based on historical recordings of landslides, available in the national landslide database (NLD). Events from the NLD have been used to perform statistical analysis of hydro-meteorological variables in order to establish thresholds that indicate when and where landslides are likely to occur. The thresholds are integrated into a complex EWS operated by a team of hydrologists and landslide experts. The risk evaluation is updated twice a day in the national web-portal for floods, landslides and avalanches (www.varsom.no). In case of landslide warnings, e-mails are distributed to regional emergency authorities and media.

Since 2010 the NVE has assessed the quality of hundreds of landslides from the NLD. The inventory comprises of data from several sources and varies a great deal in quality. The largest contributors are the Norwegian road and railway authorities, and the regional offices of the NVE. More recently, new methods of collecting data have become available such as Internet search engines and crowdsourcing. Search engines intercept information from online media and newspapers, while crowdsourcing is a way for the public (and experts) to share information through apps and web-tools that are designed to register natural hazards (www.regobs.no).

The quality assessment of landslides that are selected for statistical analyses uses various tools to improve and decide the quality of the recorded location and timing of the event. For the location assessment, tools include photos, aerial photos, topographic maps, text descriptions, Google street view, etc. To examine both the timing and a more precise triggering location on a steep mountain slope, distributed hydro-meteorological observations and simulations have also been used.

The study aims to discus aspects about landslide inventory management and quality assessment in Norway. This includes ways to identify recordings with poor quality (doubtful or human-triggered landslides) in large inventories.