On March 27, 1999, an 18-year-old man tried to swim across the Snake River about 1.25 miles below the Swan Falls dam, south of Kuna, Idaho. The river flow was approximately 22,000 cfs and water temperature was 49° F. Three eyewitnesses saw him swept downriver and heard him cry for help before disappearing, approximately three-quarters of the way across the river and about even with the tip of an island.

    Water flow, over the following four-week period, varied from 14,000 cfs to 22,000 cfs; the current varied from about 4 to 7 mph; temperature remained at 49° F through the first three weeks, then rose to 54° F. The Ada County Sheriff's Department handled the search which was coordinated by Lt. Raney and Sgt. Roberson.

    On April 3, April 11, April 20, and April 25, search dogs from Idaho Mountain Search and Rescue Unit (IMSARU), working from boats, indicated that scent was coming from a spot just a bit down-current from the place last seen (PLS). Each search used from one to four dogs and all alerted on the same down-current line. One of the major problems in working scent on fast water, of course, is the difficulty of marking just where the alerts occur. Buoys are practical on still water but not on whitewater. We had the advantage this time of using a boat piloted by Gene Ralston, a river hydrologist who volunteered his skills, his Differential GPS (DGPS) connected to a laptop computer in the boat and his sonar. Whenever a dog alerted, a tap of the key marked the spot, which could then be printed together with other alerts from the same dog, alerts from other dogs, alerts from other days. The dog handlers were able to focus entirely on their dogs, plus consulting with the boat driver and making sure the computer operator noted the alerts.

    The results were amazing. When differences in winds and water velocity were considered, all the alerts on all four search days indicated the same spot. The sonar showed an object of approximately the right size and shape; however, it was in a hole behind a boulder where grappling hooks were unsuccessful and the water current was too dangerous to risk divers.

    Red, fluorescent dye (water soluble wax form) was placed in a mesh bag and anchored at the bottom of the river, at a depth of 15 feet, near the most upstream dog alert location. The location of the anchored dye was determined using sub-meter DGPS. The dye plume was allowed to stabilize and then the location where it reached the surface was again determined by DGPS. The distance between the two locations was 75 feet. Assuming that scent would be carried in the water in a similar fashion, we used this as an estimate of the location of the scent source. Unfortunately, the most likely location was judged to be too dangerous for divers.

    On April 26, the body was spotted at about 10 a.m., floating with the current, about twelve miles downstream, which was estimated as a four-hour trip from the search site. The assumption is that the body remained in the spot indicated by dogs, DGPS and sonar for four weeks before becoming buoyant enough to move downriver.

    The IMSARU K-9 Search Team was assisted with DGPS on four other search efforts during 1999. The experience gained in 1999 will be used to refine the use of DGPS with water certified search dogs and their combined utility to search and recovery success.

     We recognize that the equipment discussed here is not generally available to most search organizations. However, the following discussion illustrates what can be accomplished and may lead to discovery of other combinations of equipment, which is more readily available. Investigations are underway to find a free or inexpensive real-time mapping program and accurate DGPS equipment to accomplish the same task.

     Gene Ralston has volunteered the use of a boat and equipment, which he uses for hydrographic surveying. The boat is equipped with a Trimble NT300D differential GPS unit and a laptop computer. In the southern Idaho area, the DGPS unit receives differential corrections from a US Coast Guard beacon in Appleton, Washington, about 300 miles away. Using the Appleton base station corrections, the real-time derived positions are accurate to 1 to 2 meters. In areas where a beacon cannot be received or when sub-meter accuracy is desired, a second GPS receiver can be set up on a known point and used to transmit differential corrections to the boat via a 35 watt radio modem.

    The GPS unit is interfaced with a laptop computer running HYPACK hydrographic surveying software. The software can display a base map of the river shoreline, islands and the position of the boat. Base maps can be derived from USGS digital line graph files. HYPACK MAX software can also display digital orthophoto quads, in addition to several other digital map formats.

    When the dog handler signals that the dog has alerted, the location and time is recorded by pressing a key, and the location is displayed on the base map. The intensity of the dog’s alert, on a scale of 1 to 5, is also noted. The map display allows the boat operator to navigate the boat back to any of the waypoints on subsequent passes with the same or other dogs. The waypoints can also be used to navigate the boat back to likely search areas for divers or other recovery activities.

    The software can also record a continuous track of the boat for use in determining the extent of the area searched. Maps can be printed showing the track of the boat and the locations of the alerts to allow comparison among different dogs. The maps are useful to evaluate the most likely location of the subject. An example of a map of the dogs’ alerts is shown in Figure 1.

    Slow, s-shaped search patterns were made with the boat proceeding upstream toward the PLS, while a single search dog and handler were onboard. The color filled dots are the alert locations of three dogs on April 3 and the open dots are locations where search dog, Mingo, alerted on April 25, the evening before the body surfaced.

    A survey grade depth sounder is also frequently used during search and recovery operations. The depth sounder, an Innerspace model 448, has an accuracy and resolution of one-tenth of a foot and prints on 8-inch paper chart in addition to sending digital depth information to the HYPACK mapping software. Chart image is an example of the image made by passing the boat over a scuba diver lying face down on a gentle slope in 25 feet of water. Even though the diver was in contact with the bottom, the depth sounder discriminated his body from the bottom. Although other objects can make similar images, the image, coupled with the dog alert information, can be a very valuable search tool.

    A new and impressive search tool has been added to our arsenal: side scan sonar. A discussion of our experience using side scan sonar and details of two deep-water searches we conducted using side scan can be found at  Side scan sonar. Links to side scan images from the Bear Lake and Lake Powell searches are included in the side scan sonar page.

Contact information:
    Charlotte Gunn, IMSARU, P.O. Box 741, Boise, ID 83701, (208) 466-8345, e-mail
    Gene Ralston, 8701 S. Five Mile Rd, Kuna, ID 83634, (208) 362-1303, e-mail
    Idaho Mountain Search and Rescue Unit