RiskScape - Landslide Tsunami Risk

An application of the RiskScape software to assess the risk to life and property from exposure to a tsunami hazard generated from submarine landslides in the Cook Straight.       

In this case study the RiskScape software was used a part of the assessment of risk along the Wellington coastline under various submarine landslide generated tsunami scenarios. The model was used with the goal of moving away from tsunami hazard modelling and progressing into modelling potential exposure of assets within a given community.

The RiskScape software was used for its ability to assess exposure of assets to various inundation models and to combine this information with vulnerability models to calculate building damage and the likelihood of injury or death to people within those buildings.  

The study set out to model three different landslide volumes (0.1, 0.3 & 1 km3) at each of the 176 cross sections identified throughout the Cook Straight Canyon. These were approximately 1 km apart. The landslides were based on simplified bathymetry and resulted in the modelling of 528 tsunami scenarios at a relatively low resolution (100m). Combined with wave heights, topography and bathymetry the Gerris solver, an open source software in fluid dynamics, was uses to produce inundation models which were a main input into the RiskScape software.

 

Figure 3 Maximum wave height for landslides of volume 1 km3 originating in different parts of the canyon. Colours range from zero (dark blue) to 10 (dark red). Size of the tsunami and resulting inundation is strongly dependent on the source location.

The study found that the nature of the source for this type of tsunami greatly affected the outcome, exposure and inundation that occurred. RiskScape modelling was strongly reliant on the exposure or inundation model inputs. The study created correlations between the volumes of each scenario landslide; the level of inundation caused, and finally the average recurrence interval for area inundated (m3).

These correlations were the final input into the inundation model. From this information RiskScape was used to model the number of buildings exposed, based on the area of coastline that was exposed in each scenario. This was correlated to the average tsunami scenario recurrence interval giving the number of buildings exposed for each proposed return period event.

RiskScape also utilised this building exposure data to model human exposure to the hazard under night and daytime scenarios. Higher occupancy during day time throughout business and industrial areas returned results of notably higher exposure to the hazard when human lives are considered.

The study also returned interesting results with exposure increasing significantly at the1:300 year submarine-generated landslide tsunami level. This increased exposure was thought to correlate with an area of high daytime use resulting in a spike of exposure at that scenario level. 

 

Example of RiskScape Modelling output

Figure 7 Average recurrence interval (ARI) in years as a function of (Top left) Size of area inundated (kilometres squared). (Top right) Number of buildings exposed (thousands). (Bottom) number of people exposed (thousands), day-time (left) and night-time (right) cenarios. Note that although a submarine-landslidegenerated tsunami can be expected on average approximately every 140 years or so it is only tsunamis with return periods of around 200 years or higher that the risk starts to become appreciable.

Because of the coarseness of the tsunami inundation modelling, this case study did not take full advantage of RiskScapes capability to quantify potential building and human losses. At this resolution RiskScape does provide an adequate assessment of the exposure to large assets such as buildings and the total number of buildings that are expected to be affected at each predicted return period event.

While the finer details are absent this information is still of use to be considered in future management strategies. Given the level of exposure highlighted in this study a closer and possibly higher resolution look is warranted at submarine-generated landslide tsunamis and the possibly consequences that they could have on Wellington infrastructure and economy.

               

Associated Models

Riskscape

The RiskScape software is designed to assist organisations and researchers with estimating asset impacts and losses from natural hazards.

Links

RiskScape - Website and Wiki page

References

Lane, EM et al.(2015): The hazard and risk of tsunami inundation due to submarine-landslide-generated tsunamis in Cook Strait Canyon. Coasts & Ports Conference 2015. 15 - 18 September 2015, Pullman Hotel, Auckland