Introduction

Residents who live along the Pacific coast line of Lincoln County, Oregon are in danger. The hazard is the seismic anticipation of an oceanic earthquake resulting in a tsunami, one of unpredictable scale. In the event that a tsunami occurs along this shoreline, the residents of Lincoln City, Newport, Waldport and all communities in between need to be prepared to evacuate to a safe location and in a timely manner. Much of the land mass within the listed communities is within the potential reach of a tsunami with the magnitude of 8 or above, according to the Tsunami and Earthquake Hazards map created by the Pacific Northwest NANOOS Visualization System team and referenced by the Oregon state government.

Roads have the potential for residents to congregate and travel via personal vehicles or public transportation. However, it is unlikely that every resident will have access to a motorized vehicle at the first signs of earthquake. Walking must be the baseline metric in finding suitable locations that are efficient in both time and space. Moreover, this region of the Oregon coast line has several estuaries located adjacent to the major communities mentioned earlier. Simply moving inland anywhere does not guarantee safety from rising water levels and so more factors must be accounted for in solving this location problem. Elevation matters as much, if not more, than distance from the shoreline. Therefore, a suitable location must have certain spatial characteristics that appropriate both distance from estuaries and the ocean and elevation above sea level. The most efficient path to get to these locations is likely the existing road network within and around the cities of Western Lincoln County. A multi-criteria evaluation was employed, using the spatial factors mentioned above along with other hydrologic features in the interest area, in order to visualize physical locations that could serve as temporary tsunami safe havens.

 

Methods

In order to employ a sound and fruitful multi-criteria evaluation I needed to define baseline values that would ensure the desired solution(s). According to the Red Cross’ Tsunami Preparedness criteria, one must be three kilometers from the ocean shore and at least thirty meters above sea level to remain safe. In addition, anyone in a tsunami risk zone must find these spatial conditions within fifteen minutes of an earthquake’s initial shock. Factors of distance, elevation and time were considered in the design of this model.

Landscape vector features (roads, rivers, streams) were acquired by the University of Oregon GIS Library server and visualized in ArcMap 10.3. Rate path distances were then generated with specified, and unique, maximum values for each feature layer. All major waterbodies (estuaries, inlets) were assigned a path distance of 1,250 meters. The British Heart Foundation provides the average walking rate for an individual of average fitness which is twelve minutes per kilometer. Using this rate and the Red Cross evacuation time, the walkable distance from hazardous water bodies was calculated, yielding a distance of 1,250 meters in the recommended fifteen-minute time period. Rivers and streams were assigned distances relative to the concept that one is larger and more powerful than the other. Rivers was given a distance of 200 meters and streams a distance of 100 meters. These distances represent how far away someone must be in order to minimize risk of injury or death. Roads, however, possess the ability to move people to safety and so this layer was assigned a distance buffer of just 50 meters. Path distance is a buffer with graduated values for each cell, a larger value corresponding to increased distance from the source feature (eg. river polygon). The largest path distance generated became the ‘study area’ because this region represents how much ground residents are able to cover in an evacuation on foot. The Evaluation will produce optimal locations in and around this area.

A digital elevation model was imported for the interest area in order to add the factor of elevation to the safety zone location model. The DEM was reclassified in regards to Red Cross elevation criterion: 30 meters or more above sea level. All values of elevation below 30 meters were reassigned the value of ‘0’ and all values of 30 meters or greater were reassigned as ‘1’. Two types of elevation space now exist in the interest area: suitable and dangerous.

A cost surface was generated via the slope values of the DEM in order to visualize and calculate efficient pathways out of the walkable zone to safety.

Standardization of all features came next. All path distance rasters were divided by the maximum value of the raster dataset in order to create a scale from 0 – 1. This simplified value scale allows for criteria weighting to occur.

Weights were assigned to establish which locations were more or less favorable for the location allocation of safe zones. The ‘study area’ defined by the walking rate of an average person was given a weight of ‘0.5’ because most walking residents will still be within this region at the end of a fifteen-minute evacuation. The DEM was weighted at ‘0.3’ establishing its significance in the decision-making process. The evacuation surface was weighted ‘0.1’, roads at ‘0.07’, streams at ‘0.02’ and rivers at ‘0.01’. The sum of weights is ‘1’ which fits the standard of each raster layer’s value scale. The weighted layers are then aggregated, cell by cell until all cells (that intersect all included path distances) have a sum value between 0 and 1. The cells valued at or between ‘0.85’ and ‘1’ were then reclassified as ‘1’ and all cells of a lower decimal value were reclassified as ‘0’. Reclassification yielded optimal safety zone sites within the ‘study area’.

 

Results

The cells colored in light yellow demarcate suitable tsunami safety zones. All locations lie within or on the cusp of the study area, adjacent to a road and at a secure elevation. However, the majority of these locations are within the study area, leading one to question their true security in the case of a intense earthquake. Optimal locations (‘1’) comprised less than ten percent of the weighted sum model dataset, emphasizing the scarcity of safe zones within the walking distance of hazardous water bodies.

Northern coastline segment, in and around Lincoln City

Central coastline segment, in and around Newport

Southern coastline segment, in and around Waldport

Fortunately, several small-scale sites were identified throughout the interest areas all along the coast line giving scattered residents regional options to reach safety in a timely manner.

 

Discussion

Locating optimal sites for tsunami evacuation around urban communities such as Newport and Lincoln City may give community members a picture of what safety looks like in time of crisis. However, it may also leave one wanting to know more. Will these sites be capable of holding hundreds of people at once? Is the terrain of identified sites flat, steep, or covered in thick vegetation? Land cover and slope information on optimal sites would enrich, and likely narrow, evacuation possibilities within the walking zone.

It is likely obvious that the walkable distance and the recommended evacuation time, provided by British Heart Foundation and Red Cross respectively, do not meet the other criterion of distance: three kilometers at minimum. I chose to favor the walking rate metric when defining the study area because it is more realistic. As stated earlier, many individuals may not have access to a motor vehicle at the time of a quake and therefore this evaluation speaks more accurately to real-life factors such as various fitness levels and vehicle ownership. A solution to this potential shortcoming is a well-orchestrated evacuation strategy, at the municipal and county scale.

Optimal safe zones found within the walkable zone could serve as the first phase of a tsunami evacuation. These various sites collection areas where residents congregate while waiting for a private or public carpooling vehicle which could then move the evacuees to a location East, beyond the three kilometer recommendation.

Some points of critique on visualization are also in order. Land cover data would have informed the true viability of an optimal site, painting a picture of its capacity and accessibility. Likewise, cluster analyses could be performed in order to more effectively visualize where groups of sites reside in relation to cities. A graduation symbolization based on site size would further enrich the story this Evaluation seeks to tell.

The MCE designed and employed to find tsunami safe zones for the residents of coastal Lincoln County is a powerful starting point. In time of disaster, this analysis may provide hope and safety to the peoples concerned and a baseline reference for public managers to design a municipal tsunami evacuation plan. Multi-criteria evaluations are as telling as the component parts that comprise them. The scale of the problem defines its own requirements. Whether three simple conditions are enough to reach a spatial decision or the problem demands that thirty-five variables with multi-value parameters be studied in order to tease out some spatial complexity, multi-criteria evaluations can shed light of knowledge where only obscurity persists.

References

“Tsunami Survival | What to Do & How to Survive.” American Red Cross. N.p., n.d. Web. 6 Nov. 2015. <http://www.redcross.org/prepare/disaster/tsunami>.

“Walks and Treks FAQs.” British Heart Foundation. Web. 6 Nov. 2015. <https://www.bhf.org.uk/get-involved/events/training-zone/walking-training-zone/walking-faqs>.