The thing about missing airplanes is that they generally leave some sort of trace, fire, smoke, smoking hole, broken trees, aluminum debris field, disturbed snow, vultures gathering - something that can be seen from farther away than a single person can be seen. That's why the search visibility in the POD Chart starts at 1 mile, it was developed for missing airplanes and the signs they leave. Gary likes to quote my sample that uses the 1mi visibility, which is probably reasonable if we're looking for the Electra or perhaps a VW. However, in this case we're talking about looking for a person who is not necessarily out in the open, and I don't think a 4 mile, or even 1 mile search visibility is reasonable. I would not use those values if I were managing an actual search for a missing person in heavy cover.
Gary on the other hand, believes he and our Navy fliers (apparently untrained in any SAR technique) can see and recognize a person on the ground under the tree cover from 4 miles away. He apparently has very good eyes.
Andrew
I've pointed out before that you are conflating "search visibility" with "scanning range" so please re-read my attempt to convince you of this at
my post here. I also
wrote this before:"2. Perhaps you missed this part of my post:
"Look at the definition of
"Scanning Range" which is the distance that a searcher "is expected to have a good
chance of spotting the search objective" so scanning range is what you have mistaken for search visibility. To make this even more clear, the definition continues,
"Scanning range can be less than but never greater than the search visibility" so
these are obviously two different things."
3. The manual states that
persons on the ground are the second most common search subjects so the CAP contemplates searching for persons and uses the POD table to plan the search for people and to assess the effectiveness of the completed search.
If you are correct that the smallest thing covered by the POD table is the size of a car, since people are the second most common object searched for, where is the correction table that would be necessary to adjust the percentages derived from the POD tables to account for the smaller object of a person. And why don't the tables include greater distances than 4 mi because you can certainly see a crashed B-52 more than 4 mi away?"
And also this:
-----------------------------
Just for you Ric, I am attaching page 74 from the CAP manual since it appears to apply to you.
"
Scanning range sometimes may be confused with search visibility..."
gl
Pages 74-75 from ref_aircrew.pdf (47.09 kB - downloaded 22 times.)
I never claimed that a person can be spotted at 4 miles, again that is your misunderstanding of "search visibility." This is a value about the clarity of the air, that you can recognize a small object, maybe a house or a building at four miles. Lambrecht reported 30 miles visibility so the you use the 4 mile search visibility column in the POD table as that is the maximum visibility listed. With a .5 mile track spacing the farthest you have to look to see a person on the ground is 1/4 of a mile, not the 4 miles you misrepresented that I was claiming. In fact, because the strip of land is so narrow on Gardner, the farthest one can be away from a person on the ground, and still be flying over the island, is less than 600 feet for fully 84% of the circuit. In fact, 39% of this donut is less than 700 feet wide and a further 45% is less than 1200 feet wide. Only the northern end of the island is a half nautical mile wide. This means that the search planes flying down the center of the strip of land would only have to search 350 feet either side of the plane (a little bit longer than a football field) for 39% of the circuit and 600 feet for 45% of the circuit. Only on the northern tip, constituting the remaining 16% of the island, would they have to search a quarter mile either side, 1519 feet.
You now claim that the POD values are only valid for searching for large objects but I quoted your words "so 1 mile is usually the max Search Visibility used,
especially if were looking for humans instead of Electras" so you used exactly the same method of using the POD table that I used and in this prior post didn't not make any other adjustment for a search for a person. Only after I posted the computation showing a high probability of detection did you then make a new post using an extrapolation method, not mentioned in the Search and Rescue Manual, to come up with an unsupported extremely low POD. I asked you to provide us with a scanned copy of an actual search planning document, the form 104a, showing that you had used this extrapolation method for a real search that you conducted and you have not provided one. I also asked you what track spacing you have used in the past when searching for persons and how you could ever arrive at a cumulative POD that would justify ever making a search for a person in the woods based on your unsupported, extremely low, extrapolated single pass POD, and you have not responded to that request either. You made a very big change in your method for determining the POD for a search for a person in your second post compared to your first post and the only thing that had changed between those two posts was that I posted my computation.
There is a big difference between "extrapolation" and "interpolation." "Interpolation" can be quite accurate while "extrapolation" rarely is.
The Search and Rescue Manual states that searching for persons is the second most common type of search yet there is no separate POD table for this type of search or any correction table to use in adjusting the published values for POD as would be necessary if your interpretation was correct, that the tables only apply to searches for downed aircraft. I have stated before that the people who drafted this manual were compelled to use conservative numbers so as not to overestimate the effectiveness of a search. So, if the values only applied to searches for aircraft and the same tables also had to be used for searching for people then, if the calculated POD was designed to apply to aircraft, then the numbers would overestimate the effectiveness of a search for a person and so would
NOT be conservative. But, if instead, they assumed the worst case, that of searching for the more difficult object to find, a person, then the tables correctly, and conservatively, predict the quality of a search for a person and underestimate the effectiveness of a search for a larger object. The is a conservative way to draft the POD tables. So which one makes more sense when drafting this table, overestimating the effectiveness of a search for a person or underestimating the the effectiveness of a search for a larger object? Which would be more conservative? Which would result in more lives being saved?
My National Search And Rescue Manual is dated 1986. You referred us to a
CAP document dated 2005. In spite of almost 20 additional years of search experience the POD table in your 2005 document is identical to the table in the 1986 manual. There is no separate POD table for searches for people nor is there a table to make an adjustment for searches for persons even though many thousands of such searches must have been made in this period. It appears that the drafters of the 2005 document were satisfied with the existing POD tables. They are also the same tables in the CAP Aircrew Reference Text (2004) excerpts of which I have attached including the definitions section. Compare "search visibility" with "scanning range." This document does have a table of distances you can expect to spot persons on the ground, 1/2 mile or less for a person in a clearing and one mile or less for a person in the open. Looking at the 1/2 mile of less situation, a track spacing of 0.5 miles means that you will pass within half of that distance, one-quarter of a mile, of every point which places a person well within the spotting distance and even better for a person in the open. This CAP document does
NOT support your statement:
"So in theory, with track spacing of .5 miles, flying at 500 ft AGL and IF you were able to see and recognize a person in the bush at a lateral distance of 1/8th of a mile (660 ft), you'd have a 1.25% POD for a single pass" in
Reply #6.
None of these documents provide any guidance for "extrapolating" to your extremely low POD for a search for a person. The National Search And Rescue Manual manual is not a document just used by a civilian agency, the CAP, but is also an official manual of the military services, Army FM 20-150, Navy NWP-19, Air Force AFM 64-2, and Coast Guard COMDTINST M16120.5. I spent 26 years in the Army, the last 7 years in the Judge Advocate General's Corps, and there is not a chance in the world that there would
not be guidance for making this type of adjustment in these manuals if such adjustment was authorized. The military is not going to leave it up to the imagination of some low ranking person planning the search for a lost person to come up with his "extrapolation" of the POD tables. And think of the all the lawyers parachuting in when a person dies because of an improperly planned search and then the evidence comes out that some low ranking person was just "winging it" by making an unauthorized extrapolation when planning the search. Get out your checkbook.
So at the end of this debate we are left with some
anecdotes from Andrew and others saying the navy guys would not be able to spot Earhart and also some anecdotes saying that they could. Ric always criticizes "anecdotal" claims. But we also have several official government manuals clearly laying out the procedure for calculating the POD and that computation , even using Andrew's numbers and method in his first post, shows a high level of probability that they would have been spotted even in tree covered terrain if they were on Gardner when it was searched by the navy planes.
I'm going to take Ric's advice to ignore anecdotes and I'm going with the official Army, Navy, Air Force, Coast Guard and Civil Air Patrol documents.gl