uss lexington search map

[Irvine John Donald]

Hi Richie;

Thanks for the great link.  I had noticed that, because this is high resolution, you can zoom down much closer and see better detail than the standard Google Earth maps.  It makes me wonder how much detail a government spy satellite can see. 

[Harry Howe, Jr.]

The Military Search reminds me of a little ditty that I learned during my military service.  It goes like this:
When in Danger or in Doubt, Run in Circles, Scream and Shout,
Raise the Flag and Shoot the Gun, and get the Word
Well Done!

[Martin X. Moleski, SJ]

i have google earth the link is to a high resolution image

It is exactly the same photo as used in Google Earth.

Putting the two side-by-side, I guess you can get a slightly better view.  I wouldn't rank it as hugely more useful than the Google earth view, which is much easier to navigate.

YMMV.

[richie conroy]

Hi Richie;

Thanks for the great link.  I had noticed that, because this is high resolution, you can zoom down much closer and see better detail than the standard Google Earth maps.  It makes me wonder how much detail a government spy satellite can see.

well if u go to bottom of page on this link http://www.cifr.it/forum64.html

he says the japaneze have photo's an maps of islands from as early as 1931 

[Chris Owens]

What interests me about the high resolution image (thanks Richie!) is that, as I browse around areas that were pretty heavily used (such as the LORAN station area), there's just about nothing visible to my untrained eye.  That gives me new perspective on just how daunting the search task is.   Looks like there's just about nothing left of the Norwich City.  As usual, every time I see a straight line or a right angle, I wonder if it's digitization artifact or man-made object...

[Irvine John Donald]

My earlier post about it being more high resolution turned out to be a function of the PC I was using and version of Google Earth. As Jeff said you can match those shots cloud for cloud.

I think Richie that you're doing a great job at digging information out of the Internet.  Please keep it up.

[Ric Gillespie]

The graphic below is convincing that the navy used a sound analytical approach at the time in 1937 and that much ground (and sea...) was covered.

I covered this in great detail in Finding Amelia.
When I look at the graphic I am appalled that so little time and so few assets were dedicated to searching the islands of the Phoenix Group given the volume of evidence the Navy had indicating that Earhart and Noonan were there and had been calling for help.  To me, the crucial failure occurred on July 10 when Capt. Friedell (USS Colorado) sent the following message to Admiral Murfin who was directing the search from Pearl Harbor:
"With completion [of] flight this afternoon, all islands Phoenix Group have been located and carefully searched for any sign of Earhart plane or inhabitants with exception Winslow reef and reef and sandbank to the northward. The charted position of these places and for several miles in vicinity was covered twice without locating them."

The logic seemed to go:
• the post-loss radio signals indicated that the plane must be on land somewhere in the Phoenix Group
• Colorado has carefully searched all islands of the Phoenix Group and found no sign of the Earhart plane or inhabitants
• Therefore, all of the post-loss radio signals must have been bogus

At that point, Colorado was released, Lexington took over, and the rest of the search was conducted in areas based entirely upon speculation and supposition.  The post-loss signals had stopped before Colorado began searching the Phoenix Group and nobody seems to have considered the possibility that the plane may have been washed off a reef or hidden in dense vegetation.  On the voyage out from California, Lexington had drawn up a plan for sweeping vast stretches of open ocean with its aircraft. Friedell's claim that the Phoenix Group could be eliminated opened the door for that plan to be put in operation - which is exactly what happened.

[Edgard Engelman]

Jeff,
Raising several points about the Bayesian analysis (which I use extensively in my own field).
This approach existed long before 1966, and was even use extensively during the 19th century.
Some of the most spectacular successes has been the cracking of the ENIGMA machine, using essentially Bayesian statistics. And I suspect that 10th Fleet also used something similar for tracking U-boats in the Atlantic, and was not suddenly discovered in 1966.
That said, these applications were so successful because one of the most important thing in Bayesian analysis, the knowledge of the prior distribution of probabilities, was extensive. Same thing with search of the Titanic, the searcher could express a 100% certainty that it was in certain part of the Atlantic if they choose a sufficiently big part of the ocean.
In the case of AE your are sure of nothing, therefore as you said, if you feed wrong data in the formulas in won't help at all, you will only have the illusion of progressing.
Another point, the Monte-Carlo analysis in one of the tools that allows making a Bayesian analysis; the other technique was solving integrals, but with the advent of computer Monte-Carlo simulation is much easier.

Edgard

[Irvine John Donald]

Hi folks. What would happen if a second method such as the Bayesian, was calculated and the results overlaid onto the familiar Monte Carlo map?  Is that possible?  Secondly, can the results of one set of calculations validate another?  Or, since we are working with one data set, would we just get the same result?  Validation then of only the data entered. Or am I barking up the wrong tree here?

[Edgard Engelman]

Irv,
As I said to Jeff, the Monte-Carlo IS the Bayesian approach

[JNev]

Irv,
As I said to Jeff, the Monte-Carlo IS the Bayesian approach

BUT it only goes so far as to put the airplane within a certain area at a certain time BEFORE what we think the end of the flight would have been -

That is why I realize it is a similar tool (same approach, yes) but only a subset of what we'd have if we extended the tool to embrace further information - like LOP, presumptions about how attractive the Phoenix group was (we think), etc.

But I understand those are not 'knowns' - and the message that AE could be in a very large box.

[Edgard Engelman]

Jeff,
The examples of real success using the bayesian search method is practically always in cases for which the users new for sure that they were on the right track but could also afford to look in a large area; money wasn't practically an issue. And by the way don't forget that it is possible that when they found the Titanic, they new exactly where to look as it is possible that Ballard was told of the results of oceanographic searches undertaken by NATO navies for other purposes who stumbled on the wreck (don't forget Ballard was/is a US Navy officer on an official misson just days before).
In the present situation, the search by TIGHAR, the best implementation of bayesian analysis could be to model the external influences of a plane landing on the reef, where TIGHAR believes it landed, and incorporate the external ellements (wind, tide, hurricanes, current, and I don't know what). But if those ellement are known only unprecisely it is doubtfull that it will be of a great help.