CONTRA COSTA COUNTY, Calif. -- When Eloise Anderson isn't working as a graphic artist for EA Engineering, Science and Technology in Lafayette, Calif., her life goes to the dogs.
As a volunteer search-dog handler for the search and rescue team of the Contra Costa County Sheriff's Department and the California Rescue Dog Association, Anderson works with a Labrador retriever who is routinely called upon to find lost children and hikers throughout Northern California.
"The dog that I use to look for lost people is called an area search dog," Anderson explained. "This means that the dog is frequently working at a distance out of sight of the handler. Handlers generally know where their dog is but there's nothing exact about it. Determining where the dog has searched is just an approximation, especially over long distances."
At first, this situation was solved by putting bells on the dog, but this provided only limited information, especially over long distances with the dog out of sight. It was then that Anderson and her co-worker, Eric R. Sheehan, a GIS analyst at EA Engineering, came up with a solution -- GPS.
"This kind of problem requires strong consideration of geographic parameters," Sheehan said. "GPS and GIS are a natural solution to the problem of knowing exactly where the dogs are and how effective they have been in their search efforts. GPS and GIS combine to provide an optimal solution when trying to quantify areas that have been searched and those that have not. In this case they solved the problem perfectly."
Anderson said that in the type of searches she performs with her dog, there is a great probability that the person being searched for is alive. "This is opposed to a search where the person being looked for is expected to be dead. In these latter cases a cadaver dog is used. Cadaver dogs work closer to their handler in a more structured search pattern, which narrows the possibility that evidence will be disturbed."
With an area search dog the animal can be far out of the range of the handler. If a person is found, the dog will pick up an orange rubber ring hanging on its neck called a "bringsel," and return to the handler. Since the dog can be a substantial distance away from the handler when she makes the "find," it may take several reports and refinds before the handler actually reaches the lost person.
Accountability of the team's efforts is another major problem corrected with the use of GPS. "Whenever a team returns from a search area the primary factor that a search coordinator must determine is how effective that dog and handler were in their search," said Anderson. "This is called probability of detection (POD). In other terms, what area was covered by the search team and, given that the missing person was in that area, what is the percent probability that the search team would have found them? Also, dogs are a lot like people. They have good days and bad days. That's a problem that is effectively observed with the use of GPS."
On October 20, during a mock search exercise, Anderson and Sheehan strapped a small, specially-designed backpack holding a Trimble GeoExplorer II to their four-legged co-worker, Twist, and tested their ideas in a demonstration to other members of the Contra Costa County Sheriff's Search and Rescue Team. "This was an opportunity to test equipment and procedures in a non-life-threatening situation," Anderson said.
Four "lost" hikers were hidden in areas of the Mt. Diablo State Park in Northern California as part of a mock search exercise. After Twist and Anderson completed each of their search assignments, Sheehan downloaded the data obtained during the search using Trimble's Office software to postprocess the data.
"Not only did we track Twist's search path, but the GeoExplorer told us that she was searching her assigned area effectively. GPS worked like a charm. The GeoExplorer gave us quality data and the Pathfinder Office software performed elegantly. It showed us where the dog had searched as well as areas where more searching may be needed."
Further, Anderson and Sheehan could process the search path data and analyze the terrain the dog had covered. The existence of accurate base data was key to their effort.
The effectiveness of a search dog's efforts depend upon how far a scent can travel. Factors such as temperature, vegetation cover and wind conditions affect a dog's ability to track a scent. A handler develops a search strategy to optimize the search coverage for the given scent conditions. The stronger the wind and lighter the vegetation cover, the greater the distance the dog can pick up scent from the missing person.
GIS allows search coordinators to buffer the search path of the dog, providing an accurate measure of the area. A user will buffer the search path a lesser distance within heavier brush, which tends to hold a scent. "With the GPS loaded inside a backpack attached to the search dog, we are able to not only collect points everywhere that the dog has been during her search, but we buffer the path to quantify areas the dog has covered with her nose as well," said Anderson. "When this data is restored into ArcView at the conclusion of the search exercise we are able to see with greater accuracy where that dog has been and what areas could be searched more thoroughly. The GeoExplorer takes much of the guesswork out of the estimation process and gives us a better estimate of POD. The POD is much less subjective thanks to the use of Trimble GPS. "
Anderson said she sees a big future for the use of spatial information through the use of GPS and GIS within this application. "I can think of a lot of instances where this might have been a big help," said Anderson. "This is really the future of search and rescue work."
and GIS, the
[ April Table of Contents]