MN MN - Brandon Swanson, 19, Marshall, 14 May 2008

New Posts

Updates are now available at Brandon's blog: http://thesearchforbrandon.blogspot.com/.

I don't have a lot of time to devote to it, but I would be happy to answer some questions about the search efforts if you have any.

Jeff Hasse
Search Manager

President, Midwest Technical Rescue Training Associates
Vice-President, Search, Rescue, and Recovery Resources of Minnesota
Director, Jon Francis Foundation
 
After reading through some of the comments on the list, I thought I'd post a few quick clarifications:

  • The search teams have indeed considered that Brandon may have fallen into an area where he cannot be readily seen. However, the area does not have any true sinkholes. Sinkholes are a part of karst (cave) geology. The only area of the state with karst geology is southeastern MN.


  • Most wells in the area are small diameter (i.e., under 6") and would not be a danger to an adult. That being said, it is possible that there are old cisterns in the area. Farmstead windmills used to pump water into a cistern for storage. (My grandparent's farm used to have a big one with rotten boards over the top.) However, this area is high-quality farmland and very little of it goes to waste. Many of the old farmsites still on the maps have been completely reclaimed for farm fields (including concrete foundations and cisterns). The teams still are on the lookout for cisterns, just in case.


  • The most likely location for Brandon to have ended up where he could not be readily seen is in one of the many small creeks, drainages ditches, or areas of tall grass (CRP land). We are systematically searching each of these within the overall search area.


Jeff
 
Many inferences have been made regarding the fact that Brandon's cell phone rang for a time before it went straight to voice mail. Unfortunately, little can be accurately inferred from this. According to one of our members who is an executive at a regional telephone company, the ring that you hear when you call someone has little to do with what is happening at the other end (the recipient's phone). It is an artificial sound that is being generated at a switch in some building, not by the recipient's phone. It is a holdover from the old analog phone days.

The way a phone handles voice-mail when it is turned off or disabled is entirely dependant on how this function is programmed at the phone carrier's computer systems. Therefore, there is a great deal of variability regarding how it works between carriers, calling plans, and even models of phones.

While there are many possibilities regarding what happened at the end of Brandon's phone call to his father (the phone may have been lost, shorted out in water, battery dead, working but Brandon wasn't able to use it, etc.), the only thing that can accurately be inferred is that Brandon wasn't able to reestablish contact with his father after that point.

Jeff
 
We have a pretty good idea what Brandon was doing during the 47-minute phone call with his father. His father wrote up something of a "transcript" of the conversation shortly after Brandon went missing.

We know that Brandon's intentions were to meet his parents at a prescribed location in the small town of Lynd (which he thought he was near).

We know that he was walking along gravel roads for the majority of the conversation. Late in the call, we know he left the road to go cross-country because he thought it would be faster to reach his destination that way.

Shortly after, Brandon uttered an expletive and the phone went dead.

The average person walks between 1 and 3 miles per hour. That puts Brandon between 3/4 mile and 2.5 miles away from his car when the phone went dead. The Yellow Medicine River is well within this range. This fact, and some initial findings of the first search dogs on the scene, caused the early search to focus their efforts on a portion of the Yellow Medicine River (which was entirely appropriate).

Unfortunately, after numerous searches, nothing has ever been found in the river (which, over this section, is more accurately described as a creek). At some point after your efforts haven't yielded results, one must consider other search scenarios or risk what experienced Search Managers call "Scenario Lock." We are very wary of this and constantly challenge our earlier assumptions.

While it was reasonably assumed early in the search that Brandon would most likely be found at the point that the cell phone went dead, this assumption has been effectively ruled out since the area has been thoroughly searched on multiple occasions. It is now more likely (nearly a certainty) that he was able to continue walking after the phone went dead.

Based on the conditions present the night Brandon went missing, the Cold Exposure Survivability Model predicts that Brandon was likely able to continue walking approximately 2.5 hours after the phone went dead (~3:10 am) before succumbing to hypothermia. The actual time is dependant on how wet Brandon got when walking through tall wet grass and whether or not he fell into water.

Based on this model and other methods we use to determine where to search, the initial overall search area was just under 100 square miles. Based on the findings the Human Remains Detection (HRD) canines, the area in question is currently approximately 120 square miles. This should give you some idea of the scale of this search.

Jeff
 
Jeff, thank you for the additional information.

I still think something happened that caused him to utter that expletive and it was not the phone going dead. What may have happened was as he was walking in the dark and talking on the phone he may have walked into some object, a post, fence, etc, that startled him and knocked the phone out of his hands. He may not have been able to find it again in the dark or if it fell into water it may have ceased working.

so if you are able to trace his initial route on the gravel road to an approximate point where he started cutting through a field, you might want to look for man-made obstructions that might have caused him to curse and drop the phone.
 
webrocket said:
I still think something happened that caused him to utter that expletive and it was not the phone going dead. What may have happened was as he was walking in the dark and talking on the phone he may have walked into some object, a post, fence, etc, that startled him and knocked the phone out of his hands. He may not have been able to find it again in the dark or if it fell into water it may have ceased working.

so if you are able to trace his initial route on the gravel road to an approximate point where he started cutting through a field, you might want to look for man-made obstructions that might have caused him to curse and drop the phone.
Click to expand...

You are absolutely correct. We have a pretty good idea where that was (because of a trailing dog used within the first couple of days). The terrain is rugged along the river with hundreds of obstructions (old fence lines, branches, animal den holes, junk, etc.) that could have tripped him up. Added to the rough terrain are the facts that the moon had just set leaving him with little light to navigate by and he naturally had poor depth perception.

I personally believe that he tripped or stumbled (uttering the expletive) and dropped the phone, rendering it inoperable. I don't know if he lost the phone or kept walking with it, but I suspect that we will find it near his remains.

The phone hasn't been found for lack of trying. This area of the river in question has been searched no less than 57 times! (Mostly before I took over the search.)

Jeff
 
Every Search Manager must deal with two fundamental questions during a missing person search: Where do we search? and How do we search? It can take decades of experience to learn how to effectively and efficiently answer these questions. I will try to give you an idea of how we go about determining where to search by first giving a quick history of our involvement in the search and then describing the process we use to establish search priorities.

SRRRMN members became involved in the search for Brandon by the third weekend of the search (June 1st). We typically try to stay out of the spotlight; so early news accounts don't mention us much. By this time the family had connected with Gary Peterson, a relative unknown in the established search and rescue community. While Mr. Peterson was in charge of much of the searching in 2008, SRRRMN members periodically searched throughout the summer and fall. News accounts chronicle the numerous attempts to involve various SAR groups in the search during this time.

In late fall of 2008, Brandon's parents asked that I take over the management of the search. The Sheriff shared copies of all the search reports that had been generated thus far. Over the winter of 2008/2009, we conducted a very lengthy deconstruction and reconstruction of the search efforts to date. We reviewed all the reports, summarized them in a Master Search History, plotted everything to a Master Map, and calculated a Probability of Detection (POD) for each and every search. This took over 300 hours of work.

In February of 2009, we convened a team of experienced Search Managers and conducted a planning meeting. There, we used four standard methods to determine our overall search area and initial search priorities:
1) Theoretical Maximum Distance Travelled (i.e., Rate of travel x time walking). This gave us the 0.75 to 2.5 mile initial planning ring.
2) Statistical. Retrospective studies based on the International Search and Rescue Incident Database found that, in these circumstances, subjects were found 5.6 miles from the Last Known Position 95% of the time, yielding a total search area of 98.5 square miles.
3) Subjective Considerations. This includes evaluating attractants, navigation aids, and navigation barriers that might influence where the subject would go. This yielded a number of "investigative clues" that helped us set initial search priorities.
4) Deductive Reasoning. We came up with and evaluated a number of scenarios that might explain the circumstances of his disappearance.

The overall search area was divided into manageable segments and then a consensus was reached regarding initial search priorities (referred to as Probabilities of Area). We then plugged each previously calculated POD into a formula that allows us to shift search priorities based on how thoroughly a segment is search (its Cumulative POD). This process yields a new list of search priorities.

The process also helps us determine what no longer needs to be searched. One of the clear conclusions of this process was that the Yellow Medicine River no longer needed to be searched. The conservative Cumulative Probability of Detection for the section of river that received so much attention was 99.997% If Brandon's body were in the river, he would have been found!

By March of 2009, we had a clear Search Plan in place. We conducted a community meeting on March 20th to explain it and it seemed to be well received.

Since then, the plan has been put into effect, and SRRRMN teams have conducted in excess of 215 missions over 25 days of searching. Certified Human Remains Detection canines have generated a large body of evidence, which has directed the search to the northwest of Porter. We believe we are very close to finding Brandon's remains.

I hope this helps readers understand why we are where we are.

Jeff
 
Hello Jeff :)

After reading about how you came to the conclusion about why you are where you are, I completely understand your search methods. I hope Brandon is found soon, so that his family can have some closure. I would like you to know that I have 100% confidence in your methods. I have nothing but complete confidence that you will return Brandon to his family. Thank you so much for keeping us updated.

I hope and pray everyday that Brandon will be brought home...

Jessica
 
When Brandon went missing, I really never dreamed it would take this long to find him. My prayers go out to those still searching for him and I sincerely hope he will be found soon and returned to his family.
 
Thank you for your support.

When a search goes on as long as this one has, there is a tremeandous amount of frustration among everyone involved, both directly and indirectly. Speculation tends to become more and more wild. The rumor mill takes on a life of its own. Conspiracy theories abound.

However, one of the best questions one can ask about a lengthy search is; "Is there a comprehensive search plan in place?" Many times the answer is no and search efforts lack direction or the same areas are searched over and over.

There are only two reasons why Brandon hasn't been found yet. 1) He is located in an area that we haven't searched yet. or 2) We searched where he is, but missed him. The former can be resolved by sticking to the search plan. (It is also possible that he is outside the area entirely, but that is extremely unlikely based on the massive amount of data generated by the HRD dogs.) The latter issue keeps me up at night. However, a careful analysis of the clues and wind patterns as well as calculating Cumulative Probabilities of Detection help direct us in this regard.

It is amazing how often inexperienced Search Managers fail to consider the former problem. In Lost Person Behavior by Robert Koester, he describes that, in the vast majority of failed searches, Search Managers did not search beyond the median statistical distance. By definition, these searches have a 50:50 chance of failing.

The main reason for this is likely "Scenario Lock" as described in an earlier post. It is common in searches for assumptions made early in a search (based on the flimsiest of evidence) to become written in stone over time. I believe it was quite reasonable early on to follow up on the presumption that Brandon fell in the Yellow Medicine River and drowned. However, one can obtain a reasonable Probability of Detection in 4-5 searches (especially when using quality HRD dogs). 57 is beyond overkill. This is Scenario Lock.

It has been a tough sell among some people to convince them that the evidence suggests that Brandon continued walking beyond the Yellow Medicine River. But the evidence clearly suggests that he did.

Like any good investigator, we follow the evidence. We can't predict when the search will be completed, only that the plan is clearly showing us what comes next. Searches seem to evolve on their own time schedule.

Jeff
 
Another quick clarification:

Several readers have expressed frustration over the lack of clues in this case. This is also pretty common in searches.

There are three types of clues that are typically found in missing person searches:

1) Investigative Clues. These are such things as the circumstances of his disappearance, his behavioral profile, the position of the moon in the sky, lights in the distance, etc. These are "soft" clues, but may collectively tell a compelling picture.

2) Physical Clues. These are footprints (now long gone), things he carried or wore, and his physical remains. It is correct to say that we have none of these in this case. However, this is not uncommon in these types of missing person cases. I suspect that most of things Brandon wore or carried are still on him or very near him. I also suspect that, once we find a single physical clue, the search will be over very shortly after.

3) Canine Clues. These are areas where certified Human Remains Detection dogs displayed behavior consistent with interest in human remains. We have over 500 points of interest in this case, which is a staggering amount. Why so many? The constant, driving winds in the area have moved scent great distances and caused the scent picture to be quite confusing.

However, both the investigative clues and the canine clues suggest that Brandon's remains will be found to the northwest of the LKP.

Jeff
 
MTRTA Search Manager said:
Every Search Manager must deal with two fundamental questions during a missing person search: Where do we search? and How do we search? It can take decades of experience to learn how to effectively and efficiently answer these questions. I will try to give you an idea of how we go about determining where to search by first giving a quick history of our involvement in the search and then describing the process we use to establish search priorities.

SRRRMN members became involved in the search for Brandon by the third weekend of the search (June 1st). We typically try to stay out of the spotlight; so early news accounts don't mention us much. By this time the family had connected with Gary Peterson, a relative unknown in the established search and rescue community. While Mr. Peterson was in charge of much of the searching in 2008, SRRRMN members periodically searched throughout the summer and fall. News accounts chronicle the numerous attempts to involve various SAR groups in the search during this time.

In late fall of 2008, Brandon's parents asked that I take over the management of the search. The Sheriff shared copies of all the search reports that had been generated thus far. Over the winter of 2008/2009, we conducted a very lengthy deconstruction and reconstruction of the search efforts to date. We reviewed all the reports, summarized them in a Master Search History, plotted everything to a Master Map, and calculated a Probability of Detection (POD) for each and every search. This took over 300 hours of work.

In February of 2009, we convened a team of experienced Search Managers and conducted a planning meeting. There, we used four standard methods to determine our overall search area and initial search priorities:
1) Theoretical Maximum Distance Travelled (i.e., Rate of travel x time walking). This gave us the 0.75 to 2.5 mile initial planning ring.
2) Statistical. Retrospective studies based on the International Search and Rescue Incident Database found that, in these circumstances, subjects were found 5.6 miles from the Last Known Position 95% of the time, yielding a total search area of 98.5 square miles.
3) Subjective Considerations. This includes evaluating attractants, navigation aids, and navigation barriers that might influence where the subject would go. This yielded a number of "investigative clues" that helped us set initial search priorities.
4) Deductive Reasoning. We came up with and evaluated a number of scenarios that might explain the circumstances of his disappearance.

The overall search area was divided into manageable segments and then a consensus was reached regarding initial search priorities (referred to as Probabilities of Area). We then plugged each previously calculated POD into a formula that allows us to shift search priorities based on how thoroughly a segment is search (its Cumulative POD). This process yields a new list of search priorities.

The process also helps us determine what no longer needs to be searched. One of the clear conclusions of this process was that the Yellow Medicine River no longer needed to be searched. The conservative Cumulative Probability of Detection for the section of river that received so much attention was 99.997% If Brandon's body were in the river, he would have been found!

By March of 2009, we had a clear Search Plan in place. We conducted a community meeting on March 20th to explain it and it seemed to be well received.

Since then, the plan has been put into effect, and SRRRMN teams have conducted in excess of 215 missions over 25 days of searching. Certified Human Remains Detection canines have generated a large body of evidence, which has directed the search to the northwest of Porter. We believe we are very close to finding Brandon's remains.

I hope this helps readers understand why we are where we are.

Jeff
Click to expand...

Jeff, Thank you for this post.
And I hope Brandon is found soon.
 
As mentioned in a previous post, the second fundamental question every Search Manager much answer is "How do we search?" This too, can be a complex problem which requires years of experience to learn how to safely, effectively, and efficiently manage.

It is very common among untrained or inexperienced Search Managers for them to simply throw people at the problem. Intuitively, this seems to make sense. While there are times when it is very appropriate to use "emergent volunteers" from the community on a search (e.g., in the early phases of a live search), historically, their widespread use poses some problems. Large numbers of searchers pose a large logistical burden on limited local resources. It is often difficult to manage and control untrained searchers (not because they are unruly, but because they do not understand the command structure or procedures). It is difficult to obtain adequate documentation from emergent volunteers. They are often not as well prepared for the rigors of searching, so they are less safe in the field as trained searchers. Untrained searchers are not as clue conscious as trained searchers. Untrained searchers do not know how to preserve their finds as evidence. The list goes on.

What is the simplest and most effective way to reduce these problems? Through the use of trained searchers. This isn't always easy either. We work hard to match the strengths of various search modalities with their search missions. For example, we try to use horse-mounted teams (Posses) in areas with a lot of tall grass, where their height gives them a distinct advantage.

Search theory tells us that three factors determine how easy an object is to find: 1) The object itself. A downed plane is obviously easier to find than a missing cell phone. 2) The environment. Heavy scrub or thick forest will hide an object more effectively than a mowed field. 3) The sensor. Different sensors (eyes, dogs' noses, FLIR, etc.) have strengths and limitations also.

In Brandon's case, we have several things working against us. First, the "object" (Brandon's body) is tough to see. Most untrained searchers believe that a body is relatively easy to spot. It isn't. Few people realize that in warm moist conditions a body can become nearly fully disarticulated (skeletonized) in as little as a week. Using the formula developed by the University of Tennessee Forensic Anthropology department, we calculated that this occurred by mid July at the latest. Fresh skeletal remains blend very well with the natural environment and are very difficult to spot. Bones can take a few years to bleach out to a point that they are easier to see.

Secondly, the environment in this case makes spotting remains difficult. The regions where Brandon is most likely to go unnoticed are areas with very dense woods, brush, or grass. The thick Canary grass in the area alone is quite effective at covering up disarticulated remains. While one would think that it would be easy to spot remains in the wide-open fields of crops, it isn't; especially if his remains have become tilled into the soil.

Thirdly, the effectiveness of a living sensor is very dependant upon its "mental map" of the object it is looking for. In other words, if one hasn't seen fresh disarticulated remains in the natural environment before, one is not likely to spot them even if looking right at them.

So what is the best way to maximize our search effectiveness given the difficult conditions of this search? Clearly, the best sensors for these conditions are the noses of Human Remains Detection dogs. Their mental maps are developed specifically for this object under these conditions. The handlers learn to interpret body language of the dog when the dog is sensing the odor of human remains. The dog then attempts to follow the scent towards the areas of greatest concentration of odor, eventually leading to the source of scent. At that point, the dog gives a trained final response.

While, as I have explained in previous posts, we have a large body of evidence that the dogs have been sensing the odor of human remains, we have yet to come to the source of that scent (Brandon's remains). This is because the constant driving winds in the area blow scent a long distance and concentrate it along scent pools (heavy grass, treelines, etc.). We have had to systematically search each of these areas which has taken a lot of time. However, with each search we narrow down the search area.

I wanted to share this explanation with you because some have expressed concern that we are no longer using emergent volunteers. I hope this helps readers to understand why.

Jeff
 
Wow thanks Jeff for all of your explanations, your hard work, and dedication in trying to locate Brandon. I wasn't at all familiar with his case until reading on CNN this morning. I'm not sure how I missed his story being posted here. It must be awful for Brandon's parents having been talking with him on the phone for that period of time and then to literally have him disappear off the face of the earth so to speak. I can't imagine what they must be going through. He was a handsome young man and I know that I speak for many here when I say that his family has our deepest thoughts and prayers while the search for their son continues. God bless you for what you are trying to accomplish.
 
Wow, Jeff! I feel like I just took Search Techniques 101 and passed with flying colors! Thanks so very much for your time and information.

To any Mods out there ... these post are so valuable and pertain to pretty much everything we look at at WS. is there somewhere, somehow that the posts can be combined and made into a sticky or put in a 'reference manual'??
 
So there's no chance he's alive? That he was kidnapped or something? When/why was that ruled out?

Was foul play ruled out?
 
marigold0909 said:
So there's no chance he's alive? That he was kidnapped or something? When/why was that ruled out?

Was foul play ruled out?
Click to expand...

His car went off the road on a deserted stretch in the middle of nowhere. He left the road and was walking through a field when his father lost communication with him. The odds that he both made it back to the road and that someone picked him up are infinitesimally small.
 
marigold0909 said:
So there's no chance he's alive? That he was kidnapped or something? When/why was that ruled out?

Was foul play ruled out?
Click to expand...

Very good question.

As a civilian Search Manager, I have to be very clear about the limitations of my role in the cases we work. The investigation of criminal activity belongs solely to the law enforcement agency having jurisdiction over the case. In Brandon's case, this is the Lincoln County Sheriff. I know he is very dedicated to this case and the department has followed up on any viable lead.

While we do conduct our own type of investigation, our goal is purely to obtain data that will help us determine where to look. We share all of our data with the law enforcement agencies to which we report, but they may not share everything they know with us. This is appropriate, since we don't need to be distracted by things that are irrelevant to our goal--finding Brandon.

This is not to say that we ignore the possibility of foul play. The possibility of abduction, hit and run scenarios, etc. are considered as we develop the probability density map. However, those scenarios generally result in the subject being outside of the search area in the "rest of the world," which is generally too open-ended for a good search plan. We explore the scenarios we can do something about.

Having said that, we do hedge our bets some. We have a standing offer to the Sheriffs that, if they have a lead that would put the subject outside our search area, we would be happy to follow-up on it. We also ask that, if they have strong evidence that the subject is not in our search area, that they share that with us so we can reevaluate our search efforts.

While I cannot speak for the Sheriffs in this case, they have publically said that there is no evidence of foul play in Brandon's case. This doesn't mean that it can be ruled out, just that--as of this moment in time--there is no evidence to support it.

Another role we play in the overall investigation of a missing person case is exclusionary. If we execute a good search plan and come up with no results, that generally reduces the possibility of a lost person scenario and increases the possibility of another scenario that puts the subject outside of the search area, such as criminal activity. There are mathematical models to assist us with this process known as shifting Probabilities of Area.

So, to summarize this case: While there is no evidence to suggest foul play, there is considerable HRD canine evidence to suggest that human remains scent is in the area. As of this date, no human remains have been recovered. While we cannot be cannot be sure that this scent is from Brandon's remains (HRD dogs are not scent-specific for specific individuals), Occam's Razor suggests that this is the most likely explanation.

Jeff
 
wadahoot said:
To any Mods out there ... these post are so valuable and pertain to pretty much everything we look at at WS. is there somewhere, somehow that the posts can be combined and made into a sticky or put in a 'reference manual'??
Click to expand...

I would be happy to participate in such an effort. I have been perusing the forums on my cases for years, but finally decided to post when it became clear to me that the general public just doesn't have access to the standards and best practices that have been around in the search and rescue community for years.

Jeff
 
You must log in or register to reply here.

DNASolves

Who is Escambia County John Doe (1989)?

Members online

... and 95 more.
Total: 2,959 (members: 96, guests: 2,863)

Online statistics

Members online
96
Guests online
2,863
Total visitors
2,959
Totals may include hidden visitors.

Forum statistics

Threads
592,286
Messages
17,966,706
Members
228,735
Latest member
dil2288
Back
Top