1 Introduction
Dogs have been successful in detecting swimmers and cadavers underwater (Eisenhauer 1971; Stanley 1981), but there does not appear to be any published information on the use of dogs to search for explosives, drugs, and other materials in water. This chapter will focus on the use of dogs to detect scent from bodies underwater although the information would also be useful when searching for other underwater sources. Much of the material in Sections 1 through 2 has been extracted from Osterkamp (2011).
Water search dog teams are used to narrow the search area for recovery operations. Te information the handler needs to provide to recovery agencies is the most likely location of the body based on the training, experience, and competence of the team. Information obtained with a water search dog team is not exact because of water depth, movement, and turbulence of the water; presence of an extended scent print (oily flm) on the surface; and movement of the scent plume by wind. What is usually provided by the team is the location of a small area likely to contain the body. Scent from the body is transported to the water surface primarily by buoyancy and from the water surface to the air by volatilization and bubble bursting. It appears that these processes create a scent print on the water surface and scent plume in the air above the water surface. Te handler’s job is to use the dog to detect the scent plume and fnd the area on the water where the scent frst comes to the surface. This information must then be interpreted by the handler to provide their best estimate of the location of the body.
Our understanding of the nature of scent, scent-bearing materials, and scent movement in water is incomplete. Consequently, it is not possible to state with certainty the nature of all scent sources, scent movement to the water surface, behavior of scent on the surface, how the scent gets into the air, and what it is that the dog smells. However, there is extensive literature on the behavior of gases, liquids, and solids in water that can be used to improve our current understanding. This chapter uses that literature to identify likely sources of scent and scent-bearing materials from submerged bodies, to suggest potential scent transport processes from bodies through the water to the air–water interface and thence into the air. Te information is used to examine hypotheses about human scent in water, scent movement, and implications for water search dog training and deployment. Realistically, it can serve only in the interim until the necessary data from research specific to submerged human bodies is available.1.1 The Body in Water
Drownings are classified as wet (water in the lungs) or dry (no water in the lungs). Dry drownings (<15% of drownings) may occur because the subject is dead on entering the water or because of the mammalian dive refex which excludes water from the lungs. This dive reflex is associated with cold water and is more common in infants <1 year old (Teather 1994).
The body in water experiences forces that cause it to weigh less than in air, change shape and density, and move about in three dimensions. One cubic foot of freshwater weighs about 62.4 lbs at 32 to 60°F (64.1 lbs for seawater) and the pressure on its bottom surface is 0.433 psi (62.4 lbs/144 in2 ) and 0.445 psi for seawater. For a column of water of depth, d (f), the pressure at its bottom is 0.433 times d (0.445 times d, for seawater). At depth, d, this pressure acts equally in all directions. There is a misconception that water pressure “holds a body down” (i.e. there is a net downward force on the body). Te above discussion indicates that the pressure at the top of a horizontal body is less (the depth is less) than at the bottom where depth is greater. This means that the body experiences a net upward pressure or force that is equal to the weight of the water it displaces. Tus, the efect of pressure on the body is to buoy it up rather than to hold it down. However, the body of a person who has just drowned is typically heavier than the water it displaces, so the body continues to sink until it reaches the bottom.
Another effect of pressure on a body is to compress it into a smaller volume, increasing its density. Te reduced body volume displaces less water which decreases its buoyancy so that body weight increases with depth. About 100 bodies were weighed underwater to depths of 100 f, and the results verifed that there was an increase in weight with depth (Teather 1994). Te experiment also showed that adults who weighed from 110 to 200 lbs weighed 7 to 16 lbs underwater. This weight was sufcient to resist movement by currents <1.5 mph (2.2 f/sec). However, water velocity decreases close to the bottom and the depth where the velocity was measured was not stated. It is assumed herein that it is the velocity at the top of the body, roughly 1 f above the bottom. Dry drownings would have air in the lungs, be more buoyant, move with less current, and may float earlier than wet drownings.
Generation of decomposition gases in the body causes its volume to increase which displaces more water and increases its buoyancy. Te expanding body may eventually attain neutral buoyancy (zero weight) and then positive buoyancy causing it to rise. As it rises, decreasing pressure allows the body to expand more which increases its buoyancy and accelerates its ascent to the surface. For the subjects in the above experiment, neutral buoyancy requires decomposition gases to expand the body volume by about 190 to 440 cubic inches (Osterkamp 2011).
As the depth increases, the pressure on the body increases so that more gas is needed to expand the body against the increased pressure and attain positive buoyancy (i.e. more gas is needed to make the body foat). Deep lakes tend to be cooler, and cool temperatures reduce decomposition and gas production which increases the time required to foat and reduces the likelihood that the body will foat. Recovery operations for drowned subjects show that some bodies in deep cool water never foat. Divers experienced in body recovery estimate this depth at <100 f where the pressure is on the order of 40 psi, experienced water search dog handlers put it at about 100 f, and theoretical calculations suggest about 180 f. Variations in water temperature and other factors that influence gas production in the body can probably account for these differences.
In lakes with currents and in streams, a body may move before neutral buoyancy has been attained. For currents <1.5 mph (2.2 f/sec) at about 1 f above the bottom, some buoyancy must be attained before the body moves, but for higher velocities the body can move without an increase in buoyancy. Observations of the movement of ice on stream bottoms weighted with sediment and almost neutrally buoyant (Osterkamp 1977) suggest that the body would be expected to bump along the bottom until buoyancy is neutral and then rise as the buoyancy increases. This is supported by observations of body damage caused by scraping and collisions with the bottom attributed to body movement by currents or wave action (Teather 1994).
Water temperature controls water density, and therefore buoyancy, but its primary effect is to influence buoyancy through decomposition. Decomposition gas production does not significantly influence buoyancy of bodies submerged for <12 hours at 60 to 65°F, <24 hours at 50 to 60°F, or <48 hours at temperatures colder than 50°F (Teather 1994). It is not possible to use these results to predict the time to float since other factors may be involved including both physical (clothing, shoes, amount of weight carried for activities, weight carried by homicide or suicide victims) and biological ones (composition of last meal or drinks, amount of body fatty tissue, decomposition stage, presence of scavengers).
1.2 Body Scent Source
The body is the source of a host of scent materials. Dogs cannot smell a body through the water, but scent and scent bearing materials from the body enter the water and rise through it to the surface and into the air to be detected. There does not appear to be information on the nature of the scent or scent-bearing materials emanating from submerged decomposing bodies. Consequently, this information must be inferred from studies of living humans, decomposing bodies in the terrestrial environment, bodies recovered by divers in recovery operations, those found foating or washed ashore, submerged pig carcasses, and water search dog training aids. VOCs from buried decomposing bodies have been identifed (Vass et al. 2008). These compounds may be soluble or insoluble in water and lighter or heavier than water. Water may modify them and accelerate or retard decomposition depending primarily on its temperature, availability of oxygen, and whether it is salty or fresh, moving or still, or difers from normal pH. The presence of scavengers and other factors can also be important (Teather 1994).
Observations are difficult to find but decomposition studies suggest that bodies may be submerged twice, once on drowning prior to bloating and once after floating when decomposition gases have been released. Some bodies never foat and some never sink, especially individuals wearing life vests or other buoyant material and infants. Te post-drowning timeline for scent sources is strongly infuenced by water temperature through its efect on decomposition. At water temperatures in the 30s°F, decomposition is so slow that internal gas production may not be enough to foat a body for weeks, if at all. At water temperatures in the 80s°F, a body may float in a day or two.
Consideration of training aids that dogs are known to detect can aid in the identification of potential scent sources from a submerged body. SD handlers know that dogs can detect submerged clothing and shoes, possibly from VOCs in the items (e.g. from sweat, secretions) as a result of contact with the skin. Handlers also use human hair that has glandular secretions on it as a training aid. Fingerprints consist of water soluble compounds and insoluble compounds modifed by hydrolysis and bacterial degradation that include VOCs that have been implicated in human scent (Ramotowski 2001). A single fngerprint on a slide immersed in water can produce an oil flm on the water surface within a few minutes (Pearsall and Verbruggen 1982), and hair submerged in calm water quickly produces an oil flm on the surface (Osterkamp personal observations). Human bones produce VOCs and have been used as training aids for end stage decomposed bodies in water. In a study of submerged pig carcasses, the odor of bones with greasy decomposed tissue has been noted (Anderson and Hobischak 2004).This discussion of potential scent sources from submerged bodies starts with the time of drowning. For searches conducted soon afer the event, dissolution of VOCs from the skin (sweat) and insoluble oily residues from secretions would be important. Vomitus, feces, urine, and existing intestinal gases purged as a result of muscle relaxation have been implicated (Teather 1994). Injuries to skin, tissue, and bones (from homicides, suicides, propellers, body movement as a result of currents or wave action, and scavengers) can produce body fuids and particles of skin, tissue, and bones. Studies of pig carcasses show that scavengers may cause body fuids and particles to be released throughout the period of submergence especially when the body rests on the bottom. Blood, foam, and blood-stained foam have been observed coming from the mouths and nostrils of recent drowning victims, apparently due to lung damage caused by agonal gasps during drowning. A bubbly, malodorous, brownish green, blood-stained fluid has been observed coming from the mouths of drowning victims as a result of pulmonary autolysis (Teather 1994).
Microorganisms associated with putrefaction convert soft tissue to simple molecules, gases, and liquids. Gases produced in the bowels and other parts of the body during putrefaction include hydrogen sulfde, carbon dioxide, methane, ammonia, sulfur dioxide, and hydrogen (Vass 2001). Tissues are converted to volatile fatty acids and other compounds including putrescene and cadaverine that have been used to train cadaver dogs. Gases and fuids in the intestines and lungs purge from the mouth, nostrils, and rectum. Accumulation of decomposition gases in body cavities and in sof tissues lead to fotation. The remains float until they lose their putrefaction produced buoyancy and then sink.
Shedding of hair and nails and skin sloughing occurs. Saponifcation (formation of adipocere, a malodorous, cheesy, compound of fatty acids) retards decomposition. It can persist for long times, particularly where the body is covered (clothes, dive suit). A diver recovered from a depth of 886 f afer 10 years had skeletonized hands and head that were exposed to the water, but the body inside the dive suit had saponifed, appeared almost fresh, and produced a strong smell (Zimmermann 2006). Internal organs may remain in a semiliquid state. Eventually the body will be skeletonized with the remains ofen partially covered by clothes and fesh and partially held together by greasy connective tissue. Disarticulation follows with the potential for separation of bones by currents, wave action, and scavengers (Teather 1994).
The above limited survey of the fate of submerged human remains indicates that scent sources from submerged bodies include gases (dissolved and in bubbles), liquids (plumes and droplets of body fuids, secretions, and decomposition fluids), and solids (particles of skin, tissue, bones, feces, vomitus) that may be encountered in a search for the remains. These materials have been shown to consist of or have associated VOCs that can be detected by search dogs. It seems clear that the training of a water search dog should include scent sources covering the full range of decomposition and fate of these remains (Table 1). It also appears that exposing a dog to remains at the proper stage of decomposition before a search (prescenting) may improve their detection capability. Common deployment times for water search dogs range from a few hours to a few days afer a known drowning and at random and sometimes much longer times for missing persons, homicides, and suicides. There are observations of the successful use of dogs many years afer the occurrence of a drowning (Graham and Graham 1987).
Table 1 Potential Scent Sources for Training Water Search Dogs Based on the Decomposition Stage of the Body
1.3 Scent Movement in Water
The above considerations indicate that the body produces scent materials in all the common phases of matter (gases, liquids, and solids) and the physical characteristics of these materials, including solubility, phase, and density (buoyancy) suggest specific scent transport processes. Local hydrodynamic conditions, especially currents and turbulence in rivers, also play a role. Studies of hydrocarbon seeps in the seabed, gas transfer at air–water interfaces and ice formation in freezing streams provide insight into the characteristics of these processes. Potential transport processes available to move scent from the body to the water surface and into the air above the water are examined below.
1.3.1 Gases
There do not appear to be any measurements or observations of the gases emanating from a submerged decomposing body, but it is likely that these would be much the same as those from a decomposing body in the terrestrial environment. Decomposition gases, foam, and bubbly fuids observed coming from the mouths and nostrils of bodies, and bubbles released from clothing, are examples. Since decomposition gases are soluble in water (Gill-King 1997), potential gas sources are dissolved VOCs and gas bubbles.
Potential scent transport processes for dissolved VOCs to the water surface include molecular diffusion, vertical turbulent diffusion, and entrainment in an upward flow of bubbles, buoyant liquids, and solids. Molecular difusion is too slow which leaves turbulent difusion and entrainment. When the dissolved gases reach the surface, volatilization is the most likely pathway for scent transport into the air above the water surface (Cheng et al. 2005). Investigations of submerged hydrocarbon seeps (<230 f. water depth) provide some insight into the nature of gas bubbles and their transport to the water surface. Te seeps release gases as bubbles that may be oil coated as well as oil droplets that rise to the surface because of their buoyancy and local water conditions. Te gases released are primarily methane but also include carbon dioxide and trace gases such as hydrogen sulfde, some of the same gases produced during decomposition. As methane bubbles rise, they exchange gases with the surrounding water, dissolve as methane outflows, grow as dissolved air (nitrogen and oxygen) inflows, and expand due to decreasing water pressure. Bubble rise velocities typically range up to ~1 f/sec for large bubbles (MacDonald et al. 2002).
When gas bubbles reach the surface, bursting occurs. Upon bursting, the bubbles leave an oil sheen on the water which indicates that they contained oil. Bursting gas bubbles can eject bubble contents (gases and water droplets with oil from the inside surface of the bubbles) into the air to a height of ~1 f above the water surface (MacIntyre 1974) as shown in Figure 1. Breaking waves, splashing, and wind spray can enhance gas transport from the water into the air.
The above studies suggest that decomposition gases and some fluids may be transported to the water surface by bubbles. Bubble bursting at the surface, enhanced by breaking waves, splashing, and wind spray, would eject gases and water droplets with fluids into the air above the water surface and leave a film of volatile fluids on the water surface. It would be possible for dogs to detect the gases in the air and VOCs from the films as long as gases are available from the decomposing body.
Figure 1 When a bubble containing scent reaches the surface, the film on top in contact with the air bursts and creates tiny film drops (a). Water surface forces act like a slingshot that ejects a jet of water and scent into the air (b). The jet breaks into drops that also contain scent from the inside surface of the bubble. The drops frequently evaporate, leaving their dissolved scent in the air.
1.3.2 Liquids
Urine and blood consist of organic and inorganic solutes dissolved in water with some volatile compounds and platelets indicating they are heavier than water and would sink in still water. These considerations suggest that urine and blood may not be significant scent sources except in turbulent streams and when entrained in a buoyant flow.
Studies of submerged seeps give some insight into transport processes for buoyant, insoluble body and decomposition fluids to the water surface and for their volatile components into the atmosphere just above the surface. Oil released from seafoor vents breaks into plumes and droplets that rise to the surface. Droplet rise velocities are much slower than bubbles unless entrained in an upwelling with gas bubbles. As oil droplets rise, the more volatile components dissolve in the water, and on reaching the surface the oil spreads in a thin film.
Skin secretions, body and decomposition fluids, and other fluids (from the lungs and gastrointestinal tract, broken skin blisters, other skin ruptures, greasy bones, and other remains) are generally lighter than water. Te seep studies indicate that oily fuid plumes and droplets would rise to the surface because of buoyancy and spread there in a thin flm. Dissolved gases may rise to the surface by turbulent transport and by entrainment when in close proximity to the buoyant droplets and plumes. Scent can be transported into the air by gas bubble bursting, volatilization, breaking waves, wind spray, and splashing at the surface.
1.3.3 Solids
Particles of vomitus, feces, skin rafts, skin, bone, and tissue may be transported to the surface because of their buoyancy and turbulent diffusion. These particles typically have secretions, bacteria, and various body fuids on them that produce VOCs. Dissolution of the VOCs in the water, transport to the surface by entrainment and turbulence, and volatilization at the surface would create a gas fux into the air. Some of the secretions and body fuids may be transported to the surface as plumes or droplets and into the air by volatilization. When they reach the water surface, the associated volatile components on the surfaces of the particles can be transported into the air by volatilization. Larger particles have faster rise velocities because of their larger buoyancy but the rise velocity also depends on the shape of the particles. For particles with a disc-like shape, about 0.04 to 0.2 in in diameter and density about that of ice, the rise velocities would be expected to range from0.1 to 0.8 in/sec, and for skin rafs, <0.04 in/sec (Gosink and Osterkamp 1983). Transport of volatiles into the air would be enhanced by surface water dynamics (breaking waves, wind spray, splashing).
In summary, buoyancy is the primary method for transport of scent materials to the water surface with gas bubbles rising at relatively fast rates and other materials rising more slowly. Entrainment in an upward buoyant flow can enhance the relatively slow rate of less buoyant materials. Turbulence may be efective when present in streams and rivers and may bring materials that are slightly heavier than water to the surface. Volatilization from oily flms on the water surface and from solids foating there puts VOCs into the air, and VOCs in the gas phase can be ejected into the air when bubbles burst. Wind spray and breaking waves enhance these processes. An airborne scent plume would develop from these sources.
2 Use of Dogs for Water Searches
2.1 Scent Displacement
Buoyant scent materials rise through the water to the surface where the emergent materials generate a scent plume primarily by volatilization and bubble bursting. Dogs detect the plume in the air above the water surface afer it has been subjected to prevailing winds and atmospheric conditions. Search teams (dog, handler, and boat pilot) attempt to detect and follow the scent plume to the area where the scent materials emerge from the water. Te question remains about the body location in the water. In lakes with no current, buoyant scent materials can be expected to rise vertically from the body to the water surface (Osterkamp 2011) so the body should be directly under where the scent emerges (Figure 2).
In lakes with through flow, rivers, and tidal areas, scent materials would be carried some distance downstream before they reach the water surface. Te distance carried depends on the turbulence, current velocity, and water depth (Osterkamp 2011). Laminar flow (current <2 mph, relatively smooth channels, no obstructions in the flow and a smooth water surface) is not common in rivers but can be found in lakes with through flow and tidal areas close to the time that the tide reverses. It would typically carry scent material some distance downstream underwater before the scent emerges for even shallow water depths. Irregularities in the banks and river bed, objects protruding into the flow, flow velocities >3 mph and variable channel width commonly produce turbulent flow adjacent to shore, in the main channel, behind obstructions, and downstream from dams. Turbulent flow is visible as swirls, eddies, bubbles in the water, and a rough water surface. For these common flow conditions, scent is carried to the surface primarily by turbulence and reaches the surface much closer to the source than for laminar flow (Figure 3).
Figure 2 A schematic profile of buoyant scent materials rising in a lake. These buoyant materials rise vertically from the source to the water surface through the thermocline. Bursting bubbles eject scent into the air. The bubbles and other scent materials produce an oily film that would spread on the surface and volatilize into the air.
Figure 3 A schematic illustration of scent movement in turbulent flow (rivers, tidal areas with incoming or outgoing tides). The bubbles and other scent materials produce an oily film that spreads on the surface downstream and downwind and volatilizes into the air.
2.2 Thermoclines
A thermocline is a thin horizontal zone of water found in a lake during the summer stratification that separates warm well-mixed water near the surface from cooler stagnant water below. It was formerly believed that scent from a submerged body on a lake bottom was transported to the surface by diffusion and therefore could not penetrate the thermocline (Hardy 1992). This belief led to the idea that it was best to search for submerged bodies when a thermocline was not present (winter and part of fall and spring) and was also used to explain failures of dogs to detect bodies in the presence of thermoclines. However, the transport processes for gases in bubbles, liquid plumes, droplets of oily secretions, and buoyant particulates are driven by buoyancy (Osterkamp 2011), which would readily transport scent-bearing material through the thermocline to the surface (Figure 2). Problems with the performance of water search dogs where thermoclines exist must be a result of other factors.
2.3 Scent Pooling at the Surface
There is anecdotal evidence that if the air temperature is colder (35°F or below) than the water, scent pools at the water surface and does not get airborne, which requires dogs to swim or get their noses close to the water surface to detect it (Hardy 1992). However, air in contact with the water under these conditions would be warmed by the water, making it lighter and unstable and cause convection in the air above the water surface. Te convective layer would mix air, water vapor, and scent from the water surface into the atmosphere (this can sometimes be observed as a layer of fog with a thickness of several feet or more). Also, gas bubbles that burst at the water surface eject their contents (water droplets with oily scent and gas) into the air with water droplets reaching heights of about 1 f. Wind (even a light breeze) would enhance volatilization and further mix scent into air near the water surface. Tus, the thickness of the scent plume above the water surface is likely to be on the order of feet or more which would make it possible for a dog to detect it from a boat under the above conditions (Osterkamp 2011).
2.4 Natural Decomposition Gases
Methane, hydrogen sulfide, and carbon dioxide are decomposition gases that can also be produced by decaying organic material in wet environments like lakes, swamps, landfills, septic tanks, grainfields, and outdoor toilets. CDs often show an interest in them and even alert on them. These gases are often called “swamp gas”. There are examples of lakes drained because of alerts by CDs on swamp gas. It requires considerable time, money, and energy to drain a lake. When nothing is found, it detracts from the reputation of CD teams and uses resources that could have been used elsewhere. Consequently, it is desirable to train dogs not to alert on gases from these areas. This can be done by training CDs in these areas, ignoring any interests and alerts, and rewarding for cadaver sources only. Landfills, septic tanks, grainfields, and outdoor toilets are common in rural areas and ponds and lakes with shallow water over mud bottoms and organic materials produce swamp gas and these places can be used for training.
3 Tactics for Water Searches
3.1 Introduction
Water searches bring new challenges to canine teams. Searches for drowned subjects are different from other types of searches because the dog’s movements are usually restricted to a boat. The dog cannot choose how it moves to detect a scent plume and follow it to the source and must rely on the handler and boat pilot. They become an integral part of the scenting team and are directly involved in locating the source. The handler must continually read the dog for information on the absence or presence of scent, the direction the scent is moving, and communicate this information to the pilot who is responsible for driving the boat. It is easier and more efficient when the handler and pilot are used to working together.
Usually the dog cannot see or contact the body so it must perform its alert based on the presence of scent only. There are also new and unfamiliar hazards that the handler and pilot must be able to recognize such as the effects of waves, currents, hidden obstructions, sweepers, eddies, low head dams, tides, presence of an ice cover downstream of the search area, and more. The pilot, handler, and dog should always wear life vests when working near and on water. It is recommended that the handler take a water safety class before beginning water search work.
An examination of the scent materials from a decomposing body and movement of these materials to the surface indicates the scent plume consists of decomposition gases from bubble bursting and gases volatilized from an oily film on the water surface and from any solids floating there. The primary sources appear to be the oily film and airborne scent (Figures 2, 3, and 4) that are produced as a result of bubble bursting, oily secretions from the body, and oily decomposition material. Bubble bursting is only important as long as gases are available from the decomposing body. The oily film varies in composition depending on the decomposition stage of the body.
The concept of a “scent print” is introduced herein to describe the high concentration of scent on the water surface from a drowned subject or body part in analogy with observations over buried explosives. This scent print must begin where bubbles from the body gases first reach the water surface with additional oils added as other scent materials reach the surface over the body in lakes or further downstream in rivers. It extends downwind and/ or downstream where the material is carried by wind and/or current while emitting scent into the air by volatilization and other processes (Figure 4). Volatilization of the oily film which spreads on the surface and fragmenting by surface turbulence and waves is expected to reduce the film thickness to the point where the scent print becomes discontinuous and eventually disappears downstream and/or downwind. The oily film from secretions and gas bubble bursting may be very thin so it would disappear quickly, implying the scent print may be relatively short while the film from fatty decomposition fluids may be thicker and last longer. The location of the scent print in rivers depends on the rise velocity of the scent material, flow velocity and depth, and whether the flow is laminar or turbulent as noted above. There is support for the presence of a surface scent print since oily films have been observed on the surface of calm water over a cadaver source and in laboratory settings. This scent print and airborne scent from bubble bursting are thought to be the sources of the scent plume sought by the team. While the presence of the scent print is very likely, there are no direct observations of it from submerged bodies and no information on its characteristics (length, width, thickness, volatilization rate of the film, presence of insoluble materials, etc.). This lack of information contributes to the difficulty of locating bodies in water.
Figure 4 Plan views of surface scent prints and scent plumes on a lake with no wind (circle) and a lake and river with wind. In a lake, the body is located under the upwind end of the scent print, and in a river it is located upstream of the upstream end of the scent print.
3.2 Search Tactics
There are several search methods and patterns that are commonly used in water searches with dogs which are primarily based on the type of search, wind direction, and current, if any (Bryson 1984; Graham and Graham 1985; Hardy 1992; Barton and Clemmo 1997; Koenig 2000). Usually, they involve gridding across wind and moving each grid line upwind with the dog always on the upwind side of the boat. Te boat should turn into the wind at the end of each grid line and the dog should switch to the upwind side. Once the scent plume is detected as indicated by the dog’s alert or CB, the patterns diverge. One pattern (Figure 5) is to continue gridding the plume but with shorter grid lines noting the position of each alert as the dog passes through the plume. Eventually, the boat will pass upwind of the scent print with no alert by the dog. On lakes, the body is between the last two grid lines and roughly on a line plotted through the alerts. A fne grid through that area (Figure 6) can help to locate the emergent scent print more precisely and the body will likely be under the upwind side of the scent print. If the gridlines are advanced downwind (not recommended), when the first alert occurs the body is upwind between that gridline and the previous one and a fine grid upwind of the alert as above should help to locate it.
Figure 6 shows how to locate the upwind end of the scent print more precisely. Divide the area between the alert and no alert grid lines in half and pilot the boat along this line. An alert indicates it is upwind of that line and no alert indicates it is downwind of that line. Next, divide the appropriate half of the area in half again and pilot the boat along this line. Repeat until satisfed (as indicated by prior discussion with the IC) with the precision of the results. Te independent use of a second dog can help to defne the upstream end of the scent print. The handlers must then estimate its location using the dog’s behavior, wind, current, water depth, and other factors (location of CB, alerts, and boat paths) as noted previously.
Figure 5 Plan view of a lake search with wind illustrating gridding with the first method which is to continue gridding with shortened grid lines after scent is detected.
Figure 6 Gridding to locate the upwind end of the scent print. The lower and upper lines are part of the initial grid. No alert on a line indicates the source is downwind and an alert indicates it is upwind putting it between the two lines. This distance can be quickly reduced by making repeated passes half-way between the alert and no alert lines.
The second pattern is to mimic the natural tendency of the dog to quarter the wind (grid across wind) to detect the scent plume and to quarter upwind in and out of the scent plume to locate the source (Scent and Wind). Close to the source, dogs turn into the wind and follow the plume upwind to the source. Dogs seem to know when to do this but the handler may not. When the dog first alerts or indicates it is in scent, the handler should have the boat turn sharply, about 30° to 45°, into the wind (quarter upwind) and follow the scent plume by watching the dog’s head for direction (Figure 7). If the dog indicates it is out of scent, the handler should turn back into the wind at about 30° to 45° upwind and attempt to reacquire the scent plume. Te procedure is repeated until the dog no longer detects scent or the dog leads the boat directly upwind and then alerts or is obviously out of scent. A fine grid downwind of the position where the dog no longer detected scent (Figure 6) can then be used to better define the scent print. If the decision is made to grid downwind, not recommended but sometimes necessary, once the dog alerts on the first encounter with scent, the handler should turn back into the wind and proceed as before.
A third pattern is to turn the boat into the wind when the dog first detects scent and attempt to follow the plume to the upwind end of the scent print. If the plume is lost it can be reacquired by gridding upwind from the point where the dog last had scent. While this can work well for an experienced handler and boat pilot used to communicating with each other, many handlers must work with a new pilot each time they search which can be difficult.
Figure 7 Plan view of a lake search with wind illustrating the second pattern which is to quarter upwind after scent is detected.
The patterns increase with difficulty and efficiency from the first to the third; however, on searches there is often little difference between them. Perfect grid patterns are easier to make on paper than in a boat on a lake or in a current. Do not worry about this; the object of gridding is to put the dog downwind and within scenting range of all the parts of your search area.
There are no guidelines on boat speed although most teams work very slowly, perhaps too slowly, especially for large area searches. It is recommended to move at about the speed the dog naturally uses when searching an open fled; although some dogs move fast, and a slower speed may give the handler and pilot more time to react to the dog’s CB or alert.
Observations of water search dogs during training and searches suggests that some dogs can discern the upwind edge of the scent print which appears to be the area of greatest scent intensity. These dogs show that they are obviously in scent when gridding but do not give their alert until a point where they become much more animated and may try to jump into the water. Possibly, the additional scent put into the air by bubbles bursting stimulates the dogs to alert. Dogs that are excitable, young, novices, or inadequately trained may offer their alerts immediately on encountering the scent plume which is a training problem.
In all boat searches, a handheld or boat -mounted GPS should be used to record the path of the boat and to mark whether the dog is in or out of scent. Using more than one competent K9 team is desirable to confrm the alerts and to further defne the scent print. Te position of the body must be estimated taking into account the dog’s behavior, wind, water depth, and other factors (location of CB, alerts, and boat paths). These quantities should be plotted on a map of the search area and used to estimate the location of the body. Experience in training and searching under similar conditions and discussions with local recovery units and bystanders about any past finds in the search area help to locate the body.
3.3 Shoreline Searches
Water searches for bodies near a shoreline are possible when the wind is at least partially onshore. For these searches, the wind direction and channeling by the shoreline are important. Wind blowing directly onshore is the simplest case since the body must be offshore and directly upwind. Dogs may alert at the shoreline or face offshore while staring or whining and some dogs may swim following the plume part way or all the way to the scent print. Wind blowing at an angle to the shoreline is a more difficult case since when the dog offers a CB or alerts, the handler must then carefully determine the wind direction because it is the direction to the body. This is something that is best learned during training. If the shoreline consists of a bank with a sharp change in elevation (even just a few feet) or a sharp change in vegetation (common conditions), wind and scent may be channeled along them making the problem more difficult (Figure 8). These changes may also cause scent to collect at the water’s edge or on shoreline vegetation. If dogs are not trained to follow scent upwind past scent collectors, they may alert on them.
Figure 8 Lake shoreline search with wind blowing at an angle onshore. Scent collects on anything above the water surface (e.g. stump) and channels along shore where it collects on vegetation.
Dogs can search a limited distance offshore by wading or swimming when the wind is offshore. However, this method should not be used when offshore winds are strong, when there are strong currents, or an outgoing tide since dogs have difficulty swimming against a current of several mph or strong winds. Dogs can be trapped and drowned in rivers by strainers, sweepers, log jams, and the presence of a downstream ice cover. A few advanced handlers train their dogs to take hand signals for direction while swimming offshore which allows these teams to search a hundred yards or more offshore under safe conditions. This method is especially efective in small bodies of water. The dog’s alert may consist of swimming in circles, twisting and turning in the scent, looking down into the water, biting the water, or allowing surface water to flow into their mouths and out the side.
The presence of banks, bluffs, and vegetation on land may create turbulent eddies (Above-Ground Searches) in the air near-shore that bring scent a short distance shoreward when the winds are offshore. Lake, pond, and river breezes are like sea breezes but form on lakes, ponds, and rivers during the day and offer the possibility of searching near-shore water from the land. Land breezes from sunset to sunrise make it possible to search difficult shorelines from a boat moving along the shore. These breezes are subject to the prevailing winds that may enhance or destroy them.
3.4 Boat Searches
3.4.1 Lakes
On lakes with no through flow and little wind, lateral spreading as the scent material rises is expected to be small (Figure 2). The initial surface scent print should be larger than the size of the body and increase slightly in width on the surface as the depth increases. The oily scent material on the surface would be expected to spread in a thin film and significantly increase the size of the scent print. Wind would elongate the surface scent print downwind (Figure 4), and VOCs from bubble bursting and volatilization would produce a scent plume that moves downwind. The source is always directly below the scent print when the wind is calm and under the upwind end of the scent print when wind is present.
Search tactics on lakes under calm conditions (Figure 4) are to grid the search area closely (a few tens of yards or less), and it may be better to wait for wind if the search area is large. If the lake surface is smooth, it does not necessarily mean there is no wind. So be sure to check for wind with a wet finger, surveyors tape, powder puff, or wind gauge. When wind is present (almost always), initial grid lines should be perpendicular to the wind (Figure 5 and 7). Grid line spacing before the dog has an alert depends on conditions. A rough estimate of grid line spacing with light wind would be about 50 to 150 yards. The handler needs to determine the effects of wind, decomposition stage, and water temperature on grid spacing during training. For a given wind and water temperature, searching for bones associated with an old drowning will be much different from searching for a body when decomposition is at a maximum.
Water turbulence may increase close to shore due to the presence of an irregular shoreline or in shallow water with obstructions. If the shore has a high bank, bluff, or tall vegetation, turbulent eddies may form in the air near the shore as noted for these features in Above-Ground Searches. The presence of hills, bluffs, and tall forests on the shore can modify the wind some distance offshore and in nearby bays. Lakes and reservoirs with through flow appear to behave like slowly flowing rivers, possibly with laminar flow.
3.4.2 Creeks and Rivers
Small creeks can often be searched from their shores, but those with shores that are high and steep, covered by dense vegetation or rocks, usually need to be searched with boats. These creeks often have strainers (e.g. brush piles, fallen trees) in the flow that can trap bodies, typically on their upstream sides, and sweepers (trees growing almost horizontally from the banks with limbs partially in the water) that can trap floating bodies. Some creeks have an alternating rapid, pool flow so that scent from a body trapped in rapids upstream or at the head of a pool may be carried downstream below the pool which can be searched by working the dog in rifles at the downstream end. Mountain creeks and those with very steep gradients are dangerous to search. Drainage ditches and canals, especially those that flow only at certain times, often have smooth grass covered beds and banks that favor movement of bodies over large distances. If there are dams, bodies may be trapped against their upstream sides. Schematic illustrations of scent movement in lakes with through flow, tidal areas, and rivers are shown in Figure 4. In tidal areas, scent can be carried one direction with the incoming tide and the other direction with an outgoing tide. The scent material always emerges downstream from the source (both directions for tidal areas). The scent print starts where the first gas bubbles reach the surface. It extends downstream an unknown distance while releasing scent into the air. The source is always upstream of the upstream end of the scent print in creeks and rivers.
Tactics to detect the scent plume and locate the surface scent print are similar for both types of flow. Rivers channel air as well as water so that air flow is frequently up or downstream, especially in rivers with high banks or dense and tall shoreline vegetation. While the point where the scent material reaches the surface does not depend on the wind direction, an upstream wind would retard the movement of the scent print downstream and a downstream wind would elongate it somewhat.
Barton and Clemmo (1997) and Koenig (2000) discuss search tactics to locate bodies when the current and wind are in the same direction and in opposite directions. For wind blowing downstream, it is recommended to grid back and forth across the river starting on the downwind side of the search area unless conditions do not allow it. The pilot can angle the boat slightly into the wind with the dog always on the upwind side of the boat which keeps the dog out of fumes from the boat motor. Once the plume is detected, use one of the search patterns described previously (Figure 5 and 7). The first pattern is described below for all river searches but either of the patterns described above can be used. Continue to grid upwind in and out of the plume (Figure 9) until a pass does not yield an alert. When this occurs, the upstream end of the scent print will be between that pass and the previous one and can be defined by a fine grid there (Figure 6). The alerts can be used to define an approximate line and the body will be roughly along this line and upstream of the upstream end of the scent print. Divers often work downstream of a tether to a boat anchored upstream of the scent print and along the line of alerts to find and recover the body. The downstream end of their search area is the upstream end of the scent print. Experience in training and searching in rivers will help handlers estimate the position of the body.
When the wind blows upstream (Figure 10), the scent print will begin downstream from the body and be carried downstream retarded somewhat by the upstream wind while releasing a decreasing amount of scent into the air. With wind and current in opposite directions, the scent plume moves from the scent print back upstream over it and the body which creates a scent plume upstream and downstream of the body. A grid (Figure 10) that starts on the downwind and upstream side of the search area would result in a line of alerts downstream in the plume, over the body, and toward the end of the scent print. The dog should give its strongest CB or alert as it passes the point where the scent is maximum. Maximum scent is likely to be the upstream end of the scent print and the dog should be trained to offer its alert there. The body would be upstream from that point and along the line of alerts. If the dog does not offer an alert there, the situation is not hopeless because divers can search along the line of alerts by working downstream of a tether to a boat anchored upstream along the line of alerts.
Figure 9 River search with wind and current in the same direction. The upstream end of the scent print is between the last two grid lines. Use the method in Figure 6 to determine its location. The source in a river is always upstream of the scent print.
Figure 10 River search with wind and current in opposite directions.
There is no information for wind that is blowing at an angle to the current. For wind blowing primarily downstream or upstream at a small angle to the current, scent materials will emerge directly downstream of the body. The plume will be carried downwind at the same angle to the current and will be wider than previous examples. As the angle increases, the plume will increase in width. The width will be defined by lines drawn downwind from the upstream and downstream ends of the scent print and increase with distance downwind. A line through the CBs or alerts will also be at about the same angle to the current. Grid upwind across the current as shown in Figures 11 and 12 or across the wind to detect the plume and define the scent print. The body will likely be directly upstream of the scent print.
Figure 11 River search with wind blowing primarily downstream at an angle to the current.
Figure 12 River search with wind blowing primarily upstream at an angle to the current.
When the wind blows directly across the river at a right angle to the current (Figure 13), scent materials would emerge downstream from the body and form a downstream scent print that would be elongated somewhat downwind. A scent plume would develop that would be as wide as the scent print is long and would move directly across the river. Gridding should start on the downwind side of the river and move up and downstream to advance the grid upwind (across the river). Once the plume is detected, continue gridding upwind in and out of the plume (Figure 13) until a pass does not yield an alert. When this occurs, the upwind side of the scent print will be between that pass and the previous one. The alert and path of the boat define an area as shown in Figure 13, and gridding this area should help to improve Source Scent print Current Wind Plume River search the position of the upwind side of the scent print. The body will likely be upstream of the upwind side of the scent print.
Figure 13 River search with wind and plume moving across the current.
For wind blowing primarily at a large angle across the current, the scent material will emerge directly downstream of the body and form a downstream scent print. The plume will be carried downwind at the same angle to the current. Grid upstream and downstream as shown in Figure 14 or upwind and across the wind to detect the plume and to locate and define the scent print as before. Again, the body will likely be directly upstream of the scent print. Water and air turbulence can be greater close to riverbanks because of the presence of shoreline vegetation, banks, bluffs, and irregularities along the shore. These features may also defect and channel wind and require closer spaced grids to detect the scent plume and define the scent print.
3.4.3 Scent Collectors and Miscellany
Scent collectors in the water include eddies, strainers, sweepers, buoys, docks, wharves, barges, bridge abutments, and other objects in the flow that can trap scent and bodies. Tose that extend above the water behave like similar structures on land and their effects on wind and scent are described in Scent and Wind. These features should be checked for scent when gridding the search area or be checked from shore, if possible. Eddies are commonly formed wherever current flows around, through, and over objects (rocks, logs, trees) in the water. Bubbly foam sometimes forms in these eddies and should also be checked for scent. Faint scent near the water surface may be enhanced by bubbles produced by the boat’s motor that break on reaching the surface, especially when shifting the engine between forward and reverse.
Rising water levels in large drainage ditches, rivers, and an incoming tide can cause backflow into connected lakes and tributaries and cause a body to be carried into them. When water levels drop, the body may be trapped in the lake or tributary.
Figure 14 River search with wind and plume moving primarily across the river.
Depressions in a stream bed are stagnant flow areas that can trap bodies until they gain sufficient buoyancy to float. Changing flow conditions and increasing current associated with rising water levels can redistribute bed material (silt, sand, gravel) into the depression covering the body and trap it there. Dogs may still be able to scent the body, but sonar scans of the bed may not show that the body is there.
4 Summary
This chapter focuses on the use of dogs to detect and locate bodies underwater and the information would also be useful when searching for other underwater sources. Dogs cannot smell a body through the water but can detect scent from the body which enters the water rises to the surface and into the air. Handlers use the dog to detect the scent plume, find the area where scent first surfaces, and provide their best estimate of the location of the body.
The body in water weighs less than in air, changes shape and density, and moves about in three dimensions. Pressure increases with depth so that the pressure on the top of a body is less than at the bottom. Tus, the effect of pressure on the body is to buoy it up rather than to “hold” it down.
Adults who weighed from 110 to 200 lbs weighed 7 to 16 lbs underwater. Decomposition gases cause the body to increase in volume which displaces more water and increases its buoyancy. The body eventually attains neutral buoyancy (zero weight) and then positive buoyancy, which causes it to rise. Decomposition gas production does not significantly influence buoyancy of bodies submerged for <12 hours at 60 to 65°F, <24 hours at 50 to 60°F, or <48 hours at temperatures colder than 50°F.
Deep lakes are cooler which reduces decomposition and gas production. At temperatures in the 30s°F, decomposition gases may not float a body for weeks, if at all. At water temperatures in the 80s°F, a body may float in a day or two.
Dogs can detect submerged clothing, shoes, human hair, and bones in addition to bodies and body parts. Scent from submerged bodies includes gases (dissolved and in bubbles), liquids (plumes and droplets of body fluids, secretions, and decomposition fluids), and solids (particles of skin, tissue, bones, feces, vomitus). Training aids should include scent sources covering the full range of decomposition and fate of human remains (Table 1). Presenting the dog on remains at the pertinent stage of decomposition before a search may improve their detection.
Buoyancy transports scent to the water surface with gas bubbles rising at relatively fast rates and other materials rising more slowly. Volatilization from oily films on the water surface and from floating solids puts scent into the air and scent can be ejected into the air when bubbles burst (Figure 1). An airborne scent plume develops from these sources. Turbulence is effective in bringing scent-bearing materials to the surface. In tidal areas, lakes with through flow, and rivers, scent materials would be carried some distance downstream before they reach the water surface (Figure 3).
Thermoclines are not a barrier for scent movement to the surface because buoyancy moves bubbles, liquid plumes, droplets of oily secretions, and buoyant particulates through the thermocline to the surface (Figure 2). Detection problems where thermoclines exist must be a result of other factors.
Air temperatures colder (35°F or below) than the water do not cause scent to pool at the water surface, and dogs do not have to swim to detect it. The thickness of the scent plume above the water surface is likely to be on the order of feet or more which makes it possible for dogs to detect it from a boat.
Methane, hydrogen sulfide, and carbon dioxide are decomposition gases that can also be produced by decaying vegetation in water bodies, septic tanks, grainfields, and outdoor toilets. Training dogs in these areas, ignoring any interests and alerts, and rewarding for cadaver sources only helps the dogs distinguish cadaver scent from these gases.
The concept of a “scent print” is introduced to describe the high concentration of scent on the water surface from a drowned subject. This scent print begins where bubbles from the body gases first reach the water surface. Liquids and particulates reach the surface later over the body in lakes or farther downstream in rivers. The scent print is the source of the scent plume sought by the team. It extends downwind and/or downstream where the scent material is carried by wind and current while emitting scent into the air. If wind is present on lakes, the body is located under the upwind end of the scent print and, in a river, it is located upstream of the upstream end of the scent print (Figure 4). When the scent plume has been detected, the next step in finding the body is to find the upwind or upstream end of the scent print.
Search tactics and patterns involve gridding across the wind and moving each grid line upwind with the dog always on the upwind side of the boat. The boat should turn into the wind at the end of each grid line and the dog should switch to the upwind side. When the scent plume is detected, three different patterns are used to locate the scent print. A fine grid through that area (Figure 6) helps to locate the emergent scent print more precisely.
Boat speed should be about the same as the dog uses naturally when searching. In all boat searches, a handheld or boat mounted GPS should be used to record the path of the boat and to mark where the dog is in or out of scent. The position of the body must be estimated considering the dog’s behavior, wind, water depth, current, and other factors (location of CB, alerts, and boat paths).
Water searches can be done alongshore when the wind is onshore. Wind blowing at an angle to the shoreline is a difficult case. If there is a bank with a sharp change in elevation or vegetation, scent can be channeled along it (Figure 8).
Dogs can search a limited distance offshore by wading or swimming when the wind is offshore. This method should not be used when offshore winds are strong, when there are strong currents, or an outgoing tide. A few advanced handlers train their dogs to take hand signals for direction while swimming offshore, which allows these teams to search a hundred yards or more offshore under safe conditions. A dog can be trapped and drowned in rivers where there are strainers, sweepers, log jams, or a downstream ice cover.
Lake, pond, and river breezes (like sea breezes) may form on lakes, ponds, and rivers, respectively, during the day and offer the possibility of searching near-shore water from the land. Land breezes from sunset to sunrise make it possible to search difficult shorelines from a boat moving along the shore.
Initial grid lines on lakes should be perpendicular to the wind. Grid line spacing depends on conditions, and with a light wind it would be about 50 to 150 yds. The handler needs to determine the effects of wind, decomposition stage, and water temperature on grid spacing during training. Searching for bones associated with a very old drowning is likely different (less scent) than searching for a recently drowned body.
Creeks that have alternating rapids and pools can be searched by working the dog in rifles at the downstream end of pools. Drainage ditches and canals with smooth grass covered beds and banks allow bodies to move large distances. If there are dams, bodies may be trapped against their upstream sides.
Search patterns are recommended for lakes with through flow, tidal areas, and rivers. These include patterns for wind and current in the same direction, opposite directions, and at angles (Figures 9 through 14). In each, the most likely location of the body is discussed.
Scent collectors in the water include eddies, strainers, sweepers, buoys, docks, wharves, barges, bridge abutments, and other objects in the flow that can trap scent and bodies. Foam sometimes forms or is trapped in eddies and should also be checked for scent.
Rising water levels in large drainage ditches, rivers, and an incoming tide can cause backflow into connected lakes and tributaries and cause a body to be carried into them. When water levels drop, the body may remain in the lake or tributary.
Depressions in a river bed are stagnant flow areas that can trap bodies, and changing flow conditions (e.g. flooding) can redistribute bed material into the depression covering the body. Dogs may still be able to scent the body, but sonar scans of the bed may not show that the body is there.
Tom Osterkamp
0 comments:
Post a Comment