Territorial landslide early warning systems (Te-LEWS)
EU7-Norway
In order to mitigate the risk of shallow landslides, a national landslide early warning system, operational from 2013, has been developed at The Norwegian Water Resources and Energy Directorate (NVE). The system was employed at regional scale to inform the public on the possible occurrence of the following type of landslides: debris flows, debris slides, debris avalanches, and slush flows. The service is nationwide and operational 24/7 and is supervised by 10 hydrologists/geologists that following a weekly rotating duty scheme. Through the system, daily warning levels are issued for all municipalities in the country. The warning period lasts from 06:00 UTC to 06:00 UTC each day. Decision making is based upon threshold levels with different probability of landslides occurrence, hydro-meteorological and real-time landslide observations, as well as landslide inventory and susceptibility maps.
​
1 The modelling
The thresholds used in the system have been derived from empirical tree-classification using 206 landslide events from different parts of the country (Colleuille et al., 2010), as a function of two variables: relative water supply of rain or snowmelt during 24h and relative soil saturation/groundwater conditions. The correlation model allows to identify 4 warning levels corresponding to different probabilities of landslide occurrence: green, very low probability; yellow, low probability; orange, high probability; red, very high probability. In case of yellow, orange or red warning levels, regional governors and infrastructure authorities are notified and orange and red warning levels are shown on national weather forecasts (Fig. 25). Most importantly, the margins for number of expected landslides and size of interested area for each warning level are very wide (Table 2).
Figure 25: Hydrometeorological hazard thresholds used in the Norvegian national LEWS (Colleuille et al., 2010).
In the last 2 years NVE has been conducting a revision and an update of the adopted thresholds, in collaboration with the Norwegian Geotechnical Institute (NGI), using statistical analysis of various hydro-meteorological data for registered and dated landslide events (Cepeda et al. 2012, NGI 2013a, NGI 2013b, Boje et al., 2014). In a first phase data from the entire country have been analyzed, but later two separate analyses were performed for Northern Norway and South-Eastern Norway respectively (Boje et al., 2014). A hydrological HBV-model (Beldring et al., 2003) has been used to combine, on a daily basis, relative water supply (rain & snowmelt) and relative soil saturation/groundwater conditions for the definition of an hydro-meteorological index (Fig. 26). In the LEWS this index is used in combination with a comprehensive expert judgment, data from other models and susceptibility maps in order to provide the basis for a daily evaluation of the warning level in each municipality of Norway. Instead of dealing with fixed geographical warning regions, daily warning levels are set for each municipality, depending on the current hydrometeorological situation, (Fig. 26). Thus, extent and position of the warning zones with different hazard levels are dynamic and may change from day to day.
Table 2: Criteria for evaluating daily hazard levels in the Norwegian national LEWS.
Figure 26: Hydrometeorological thresholds indicate landslide hazard in the regions Vest-Agder, Aust-Agder, Telemark, Buskerud, Vestfold SE Norway on 14.09.2015. B: Resultant early warning on level 2 “yellow level” issued for 70 municipalities on 14.09.2015.
Landslide susceptibility maps give an information on the spatial probability of landslides given a set of geoenvironmental factors (Varnes 1984, Guzzetti et al. 1999) and for this reasons they are combined, in the Norvegian LEWS, with an hydro-meteorological index to issue more precise forecasts. Two landslide susceptibility maps are available for Norway: one indicating initiation and runout areas for debris flows at slope scale (Fischer et al., (2012), a second one indicating susceptibility at catchment level, based upon Generalized Additive Models (GAM) statistics (Bell et al., 2014). To combine the landslide susceptibility map with the hydro-meteorological index a pixel-based approach was chosen. Therefore, the landslide susceptibility map at catchment level was converted into a 1km x 1km grid. Subsequently, both data sets were combined via a query using a combination matrix (Bell et al., 2014).
​
2 The response strategy
For the management of the LEWS employed in Norway three web tools have been implemented in collaboration with the meteorological institute, road and railway authorities and private consultants, to assist system managers and provide information to the public. The three web tools—xgeo.no, regObs.no and varsom.no—are employed to collect hydrometeorological data and quantitative prognoses used for the forecast and monitoring phases, to get real-time landslide events from field observations and to inform authorities and public about the warning levels issued.
The “xgeo.no” portal shows daily observations and forecast as well as hydrometeorological parameters and several quantitative information, such as thematic maps and time-series data in a web-GIS, within an open access webpage (http://www.xgeo.no). The maps, updated twice a day, show the conditions for the current day, as well as for a few days ahead. Some of thematic maps and time-series available date back to 1957 (Devoli et al., 2014). A landslide expert on duty (as member of a rotation team) uses the information provided by the hydro-meteorological model, the weather forecast, observations and available maps to define the warning zones and decide the warning levels to be issued for each zone. Even if the use of this web tool is reserved to experts, data is made available to the public, thanks to open data policy, through a web portal (Engeset et al. 2004). The portal (http://www.senorge.no), developed and maintained since 2008, is a map centric tool for visualization of temporal and spatial data (Barfod et al. 2013) and includes four main profiles: snow, water, weather and climate (Fig. 27).
Figure 27: Main profile at web interface portal www.senorge.no.
The second web-tool is a real-time database called “regObs.no” which means “register observations” (Ekker et al. 2013). Initially in 2010 the database was a tool for submitting and sharing snow avalanche observations (Devoli et al., 2014). Later, the database was extended to register observations related to other natural hazards such as landslides, floods and snow conditions. It was designed as a public tool supporting crowd sourcing and is currently available to the public as a website (http://www.regobs.no) and an app, also accessible through a web-service (api.nve.no). The technologies involved in the app are available in smartphones (i.e. camera, GPS, internet, data storage) in order to do large parts of hazard registration immediately “in field” within the app. The users can later access the records via the website to add more information, if needed. The database is used daily by landslide forecasters to register events reported in newspapers or from direct telephone calls from privates. Landslide experts working in the different regional offices of NVE and road authorities complete the database with field observations, which are recorded and visualized after 15 min in xgeo.no. There are two types of records: records pertaining to landslides that have already occurred (Fig. 28), records associated with landslide warning signs, like ground cracks or increased turbidity in a stream-water. The data collected are transferred into the national landslide database (http://www.skrednett.no) after a validation process.
Figure 28: Records at web interface portal www.regobs.no.
The LEWS is complemented by the web portal “varsom.no” (http://www.varsom.no). The word “varsom” in Norwegian means awareness. This tool is used to issue and distribute alert messages to both decision makers and the public when thresholds are exceeded in a certain area, thus the warning level exceeds level 1. The main goal of the web portal is to present and distribute daily warning messages (bulletins) for snow avalanches, floods, landslides and ice conditions in rivers. The portal was developed using a responsive html-code allowing the website to adjust to individual screen sizes, emphasizing “mobile first”, giving preference/priority to small screen displays (Johnsen 2013). Native apps have been developed at a later stage, and currently only an android version is available. To make the bulletin as user friendly and educational as possible, the bulletin page contains, in addition to the bulletin itself, relevant information such as: definitions of warning levels and landslide types, real-time weather radar images, maps that show hazard-related information, user feedback regarding the precision of the bulletin, educational information. This web tool provides 3 days warning levels for the different administrative regions. These warning details can be found by clicking on the link that opens the page of the region and then of the municipality. The page always features a list and a map of regions with the warning level issued (Fig. 29).
Figure 29: Warning levels at web interface portal www.varsom.no.
​
References
Bell, R., Cepeda, J., Devoli, G., (2014). Landslide susceptibility modeling at catchment level for improvement of the landslide early warning system in Norway. Proceedings of World Landslide Forum 3, 2-6 June 2014, Beijing.
Bjordal, H,, Helle, T.E. (2011). Landslides, avalanches and flooding on roads, statistical analysis. Traffic Safety, Environment and Technology Department. Report No. 5. 54 p. (in Norwegian).
Boje S, Colleuille H, Cepeda J, Devoli G (2014). Landslide thresholds at regional scale for the early warning system in Norway. in: Proceedings of World Landslide Forum 3. June 2-6, 2014, Beijing.
Cepeda J, Sandersen F, Ehlers L, Bell R, De Luca D (2012) Probabilistic estimation of thresholds for rapid soil-slides and –flows in Norway. NGI report No. 20110253-00-4-R dated 14 September 2012. Norwegian Geotechnical Institute, Oslo, Norway.
Colleuille, H., Haugen, L.E., Beldring, S. (2010) A forecast analysis tool for extreme hydrological conditions in Norway. Poster presented in Sixth world FRIEND 2010. Flow Regime and International Experiment and Network Data. Fez, Morocco.
Ekker, R., Kværne, K., Os, A., Humstad, T., Wartianen, A., Eide, V., Hansen, R.K. (2013). regObs—public database for submitting and sharing observations. In: Proceedings international snow science workshop, Grenoble-Chamonix, France, 7–11 October 2013
Engeset, R., Tveito, O.E., Mengistu, Z., Udnæs, H-C., Isaksen, K., Førland, E.J. (2004). Snow map system for Norway. In: XXIII Nordic hydrological conference, Tallin. NHP Report No48, Tartu.
Jaedicke, C., Lied, K., Kronholm, K., (2009). Integrated database for rapid mass movements in Norway. Natural Hazards and Earth System Sciences, 9: 469-479.
Johnsen, E. (2013). Modern forms of communicating avalanche danger—A Norwegian case. In: Proceedings international snow science workshop, Grenoble-Chamonix, France, 7–11 October 2013.