Finding Reese

In an effort to locate Reese, a series of analyses were conducted via ESRI ArcMap utilizing both raster and vector datasets. The specific ‘segment’ or region of the Pacific Crest Trail that which Reese was walking when she was reported as lost was first identified. An extent was created which narrowed down the search area within the Willamette National Forest. All terrain was first visualized as a Digital Elevation Model and then the ‘trailhead’ and ‘destination’ points were overlaid. The resolution was noted as approximately 32 x 32 (map units which were later defined as meters). The cell size could be found accessing the Source tab of the DEM layer’s Properties. This original resolution matters because all subsequently overlaid raster datasets will need to have identical resolution in order for a) mathematic analyses to work and b) to give the map reader an accurate visual of the land area in concern.

Contour lines are placed at 80 meter slope intervals. Space between lines displays an average elevation increase of approximately 262 feet.

 

 

Next, water bodies had to be accounted for. A vector shapefile of Oregon’s Lane County hydrology was added to the data model to give the rescue team a more complete vision of the area’s nature. However, in order for the area’s surface to be classified in its entirety (color scheme, distribution of attributes) the water bodies were reclassified into a raster format. It is unlikely that Reese traveled across lake water and so the water raster surfaces were assigned a value of ‘1’ while all other terrestrial surfaces were assigned a higher value due to their inherent slope. A ‘cost surface’ was then created, illustrating the relative difficulty of crossing any given cell, again, based on its individual slope. The ‘1’ value assigned to water was displayed in white which emphasized its exemption from surrounding terrain.

Land cover datasets would have enriched the cost surface and the accuracy it is meant to convey. Tree stand density and rock outcroppings would further discriminate the potential path that Reese may have taken. With the addition of these layers, the cell values located in either the tree or rock surfaces would receive a stronger weight once aggregation of all layers was completed.

 

Using the cost surface weighted by elevation Reese’s path was narrowed down to a path that traverses a ridge line heading Southeast through the center of the rescue quadrangle. The path proceeds to curve South and then left to the East to a low lying clearing. The least cost path is a direction of travel between the trailhead and the destination points which exerts the least amount of energy (lowest slope percentage) to traverse. If Reese has chosen the easiest path in front of her, she will have followed the spine of the Southeast hills that jaunt away from the East Cougar Lake shoreline. This route is approximately 4.3 miles in length.

Finding Reese is our only goal and so the visibility was compared between two high elevation points in close proximity of the path that was generated in the map prior. After the ‘lookout’ points were added to the model, the attribute table was inspected. Each point had two variables that defined it, the first being ‘ID’ and the second being ‘Count’. Count is the more telling value because it describes how many cells can be “seen” from the lookout cell. With no more than first instinct, I believe it is obvious to the viewer that the Eastern most elevation point has a wider view shed of Reese’s speculated route. Thus, the decommissioned fire tower residing on a hilltop to the North of the path’s major East curve ought to be reemployed for the search and rescue mission.

 

The greatest obstacle to the above analyses is the resolution of the hill shade and cost surface datasets. At 32 square meters, some steeper slopes were too abrupt for contour lines to even be discerned, with several lines on top of each other or adjacent. At a finer resolution, slopes could be better understood. Increasing the visibility of the landscape features and their variability would provide a more accurate terrain model and thus a more educated cost path to help narrow the search for Reese. Search efforts should be focused on the Northeastern ridge line and the relief to the South and Southeast zones around the path. These regions are demarcated by the lightest color values in the cost surface map above.