Print

[Heath Smith]

However, it is much more likely that they got a fix around 1623 Z, or even later, making the maximum D.R. error only 46 NM, see Landfall procedure navigation to Howland Island.

As we had discussed previously this would not make sense as they should have found Howland if they had a fix at 16:23 GMT. See attached image of 46NM DR error around Howland if that were the case. While it does make sense that he should have been able to do this at 16:23 GMT under "partly cloudy" conditions, this seems contrary to the fact that they did not make it. Was the 46NM DR error estimate created from the Northern offset theory that you suggested? If they came straight in, the error would be even smaller.

So, even if it was just dead reckoning all the way from the Ontario, then Noonan would have aimed 110 NM, at least, to the north-northwest of Howland which would ensure that they did not end up south of Howland at the point of intercept.

While this would make sense in a do-over scenario, this does not appear to be the case. FN probably did not plan for the worst case and probably had complete confidence that he could take observations. It does appear that they came straight in without an offset with every expectation of taking observations and using the direction finder even if that concept was flawed due to the lack of 500 Mhz TX capability for Itasca to get a fix on them (that she requested upon arrival) and her inability to get a fix on the Itasca using her voice frequencies. They also probably did not anticipate clouds and shadows being cast by the sun that made spotting Howland even more difficult (a bit of a nightmare scenario if you ask me).

I do believe that at 10:30 GMT they were able to spot the Ontario. Firstly this was at the end of a segment in the original flight plan, the expected to find the Ontario at this location. While the Ontario drifted 29 miles to the East, the Ontario was well within visual range from the point where they expected it to be. As they progressed along the flight line, the Ontario could have been as little as 8 miles offset from the original flight path. At that point in time they would have been 1/2 through the journey, passed the bad weather, climbing to 10,000ft for the remainder of the trip. They would have probably performed a quick check on the fuel consumption at that point to make sure whatever speed they chose for the remainder of the trip that they would have sufficient fuel to make the trip.

The attached PDF shows the 3 possible position reports at 05:19 GMT. If we assume that it was the Ontario sighted at 10:30 GMT, and that observation was made when they arrived at the position where the Ontario to be, this leaves about 1302 SM left in their journey to Howland. This suggests that they were achieving about 150 MPH ground speed which was the original plan in the first place that of course never attempted to account for head wind. So in the end of the 3 competing theories about the 5:19 GMT position reports, it does not matter much if you believe it was the Ontario that was spotted. If we add in your theoretical 110NM offset to the North, this would put the ground speed achieved around 164 MPH, with the indicated air speed at S.L. at 184 MPH. While possible it would seem that achieving the 150 MPH ground speed would be the most reasonable choice going forward.

As far as the theory about heading to Nauru goes Gary, you did not comment on my earlier Google Earth screen capture that shows they they certainly did not make a beeline from the 05:19 GMT report to Nauru (see attachment). While I agree this might have been a good idea for FN to use Nauru, I do not think this is the case unless he made the decision to do so after the 7:18 GMT report which seems unlikely. Since they were probably at 8,000ft as they headed East from the 7:18 GMT report, and the lights at Nauru were at 560ft, the range of visibility nearly touches the original flight line. It would make sense if they were slightly North of the flight line but we have no evidence that the lights were seen.

I did recall finding something on on the Internet and perhaps even the TIGHAR website about claims that AE was heard by the Nauru radio operator several times claiming she was seeing the lights of Nauru. This must have been discounted for some reason as it is not presented in the radio transmissions pages on this site. In any case, I do believe the report at 10:30 GMT "ship in sight" was just that, the Ontario and not the lights of Nauru.

Update - The report of seeing the Nauru lights was in the Waitt Institute Re-Navigation report.

[Martin X. Moleski, SJ]

I did recall finding something on on the Internet and perhaps even the TIGHAR website about claims that AE was heard by the Nauru radio operator several times claiming she was seeing the lights of Nauru. This must have been discounted for some reason as it is not presented in the radio transmissions pages on this site.

I would be happy to add it to the page that I created and maintained--but not on your say so.

Source with link?

[Heath Smith]

Martin,

I will have to dig for that later. Let me find the reference and I will send to you. Hopefully I can find some time after work. I cannot recall 100% if it was on TIGHAR or I found it elsewhere. Thanks.

[JNev]

All very good points above, Gary, thanks.

You've helped me understand the N-S constraint possibilities much better - that was something I was trying to sort out and had some belief that FN's ability to do that would not have been as great, so this sheds light.

As to the bright lights of Nauru's mines -

I don't disagree - those things can be done, and I never saw it as a fatal hazard (you can review my remark on it if you like).  I just saw it as a reason to avoid from my experience.  I've flown over cities and plenty of dark countryside at night, and some over the ocean.  As to the ocean, most of that experience was on the surface - open dark sea at night - and sometimes inland as well.  One of the worst things to encounter on the confined ICW along our coast in darkened regions is an oncoming barge being pushed by a tug with an intensely bright searchlight - far worse than I think NR16020 would have found over Nauru, for sure.  The skipper always seems to enjoy washing your own vessel down with that light - and you can guess about what that does as you're trying to share the channel with a multi-hundred foot long barge.

So, for me - I'd pass, but that is just me and I wasn't aboard with AE and FN.  But I can see your point about the fix, and how it could be managed.  I still see it far from likely, however - they wouldn't have had to 'go to' Nauru to use it as a useful fix, for one thing, and the less 'hunkering' one has to do, the better - just IMHO.  I'm also still struggling a bit with what a Nauru-deviation would have to do with the outcome - if FN was on his game and that was part of his plan, it shouldn't have impacted it.  But I think your good comments need more of my time to be fully understood, in fairness.

I'm left again wondering why, with all these marvelous certainties about what FN should have been able to do about fixing his position as you've pointed out, the flight never arrived at Howland.  If those tools work that well, then what could have failed FN that day?  Your response makes sense; unfortunately it also deepens the mystery on that count.

Thanks for your comments - I'll be studying that for a while yet - good stuff.

LTM -

[Gary LaPook]

All very good points above, Gary, thanks.



I don't disagree - those things can be done, and I never saw it as a fatal hazard (you can review my remark on it if you like).  I just saw it as a reason to avoid from my experience.  I've flown over cities and plenty of dark countryside at night, and some over the ocean.  As to the ocean, most of that experience was on the surface - open dark sea at night - and sometimes inland as well.  One of the worst things to encounter on the confined ICW along our coast in darkened regions is an oncoming barge being pushed by a tug with an intensely bright searchlight - far worse than I think NR16020 would have found over Nauru, for sure.  The skipper always seems to enjoy washing your own vessel down with that light - and you can guess about what that does as you're trying to share the channel with a multi-hundred foot long barge.

I also remember going down the ICW in Texas (the "ditch" is pretty narrow) making "one toot" passes with barges and towboats at night. The "close one eye" technique sure pays off.

gl

[Don Dollinger]

As I have said in previous posts, there was nothing in the messages received to suggest they did not know where they were. Right up to and including the "we must be on you" message they transmitted as though they knew exactly where they were.

They could not have known exactly where they were or they would not have missed Howland. 

LTM,

Don

[Irvine John Donald]

I appreciate your point Don but what I am suggesting is that there was no hint or suggestion in their radio messages to indicate that they thought they made an error in navigation. 

We know they weren't where they were supposed to be and an error of some sort may have been made.  But they were unaware an error had occured.  If they knew they made a mistake then AE could have added to a message that they were behind schedule or ahead of schedule or that they had been off course and had corrected the error.  But nothing like that.  The radio messages were just AE trying to contact Itasca and ger her fix for the last few miles into Howland. 

Its clear they missed Howland (or didn't see it or recognize it as their target destination).  But, based on radio message wording only, I think they believed they were on course.  Hence AE's transmission of "We must be on you...".  That line alone tells me that AE "thought" they were where they should be.  More of a mind set issue. 

[Heath Smith]

I did recall finding something on on the Internet and perhaps even the TIGHAR website about claims that AE was heard by the Nauru radio operator several times claiming she was seeing the lights of Nauru. This must have been discounted for some reason as it is not presented in the radio transmissions pages on this site.

I would be happy to add it to the page that I created and maintained--but not on your say so.

Source with link?

Martin,

I found the source, it was the Waitt Institute. I am not suggesting adding it to any page if TIGHAR thinks that this is not credible. I knew that I had read this somewhere but could not recall the source. It is also given in their Re-Navigation Report.

http://searchforamelia.org/position-reporting

“A Ship in Sight Ahead”
AE reported seeing “…a ship in sight ahead…” at about 1030 GMT, according to Harold J. Barnes, officer in charge of the radio station at Nauru Island who copied Earhart’s message.⁽⁷⁰⁾

In a letter from Mr. T. H. Cude, Director of Police, Nauru Island, to Dr. Francis Holbrook of Fordham University, he stated he heard AE broadcasting to Harold Barnes, Chief Wireless Operator at Nauru Island, several times between 10-11 PM that she could see the lights on Nauru Island. The lights she referred to were the flood-lights strung out along the two 1,000-foot cableways situated on top of the island to permit mining at night. ⁽⁷¹⁾

70 - Elgen Long, Amelia Earhart - The Mystery Solved, 20.
71 - Laurance Safford, Amelia’s Flight Into Yesterday, (McLean, Virginia: Paladwr Press, 2003) 31-33.

I also found a Post-Loss radio message topic posted on the forum way back in 2003 reported by a Irene Sexton that you might want to consider adding to the Post-Lost Radio page. Ric commented at the time that he thought it should be looked in to further, perhaps it was. This was not related to the lights of Nauru but it might be of interest.

http://tighar.org/Projects/Earhart/Archives/Forum/Forum_Archives/200311.txt

Update: Never mind the post-radio message. Ric later posted that it was not credible

[Martin X. Moleski, SJ]

I found the source, it was the Waitt Institute. I am not suggesting adding it to any page if TIGHAR thinks that this is not credible. I knew that I had read this somewhere but could not recall the source. It is also given in their Re-Navigation Report.

http://searchforamelia.org/position-reporting

“A Ship in Sight Ahead”
AE reported seeing “…a ship in sight ahead…” at about 1030 GMT, according to Harold J. Barnes, officer in charge of the radio station at Nauru Island who copied Earhart’s message.⁽⁷⁰⁾

That is already in the Transmission Timeline, along with a link to an article with Ric's comments on why he does not think the Cude letter (not a radio log) is a reliable souce.

Barnes was not on duty at the time the flight approached Nauru.

[Heath Smith]

I found the source, it was the Waitt Institute. I am not suggesting adding it to any page if TIGHAR thinks that this is not credible. I knew that I had read this somewhere but could not recall the source. It is also given in their Re-Navigation Report.

http://searchforamelia.org/position-reporting

“A Ship in Sight Ahead”
AE reported seeing “…a ship in sight ahead…” at about 1030 GMT, according to Harold J. Barnes, officer in charge of the radio station at Nauru Island who copied Earhart’s message.⁽⁷⁰⁾

That is already in the Transmission Timeline, along with a link to an article with Ric's comments on why he does not think the Cude letter (not a radio log) is a reliable souce.

Barnes was not on duty at the time the flight approached Nauru.

I did see the foot note after searching but I did not find the Cude's claims. Do you have the actual letter from Cude? Thanks.

[JNev]

Maybe you didn't understand how flying over Nauru would allow checking his navigation and the accuracy of his octant.  All celsetial navigation involves picking a position on the ground and calculating what altitude of a celestial body would be measured from that spot...

There are additional reasons to fly over Nauru. One is to determine the winds at their cruising altitude. You were trained to compute a wind correction angle using a wind vector diagram on your E6-B but few instructors teach the next computation, calculating the winds encountered in flight by using the same vector diagram... Using the Ontario for the visual checkpoint would not allow the same accuracy because of the uncertainty in the position of the ship.

Yet another reason why flying over Nauru was desirable is that it would provide an accurate starting point for the dead reckoning to Howland. The accuracy of dead reckoning can never be any better than the accuracy of the starting position...using the Ontario as a starting position has even more uncertainty than a celestial fix because Noonan could not know, for certain, the actual position of the ship. And, he had to be aware of the possibility of another ship being in the area...

Navigators are expected to use all information available to them for safely conducting the flight. In trans-oceanic navigation it is unusual to be able to take visual bearings on terrestrial landmarks but when they are available, navigators take advantage of them. I'll bet that any WW2 Air Force navigator that flew over the western Pacific would tell you the same thing, that he used every opportunity that presented itself to take a visual observation of islands to confirm and to improve his navigation.

So, compared to the slight cost of less than an additional 10 NM, it was well worth going over Nauru.

gl

I see your points - and these things are logical. 

But -

...flying over Nauru would allow checking his navigation and the accuracy of his octant.  All celsetial navigation involves picking a position on the ground and calculating what altitude of a celestial body would be measured from that spot..."

Shouldn't his octant have been OK after such a check at say, Lae, and perhaps again as the flight passed the Nikumanu Islands?

One [good reason] is to determine the winds at their cruising altitude.

Wouldn't that have been possible as the flight passed the Nikumanu Islands?

...another reason why flying over Nauru was desirable is that it would provide an accurate starting point for the dead reckoning to Howland...

I realize that Nauru would put them closer and therefore could reduce the potential error - good point, but again, wouldn't passage past Nikumanu have given confidence and a reliable 'starting' point for DR?

Navigators are expected to use all information available to them for safely conducting the flight. In trans-oceanic navigation it is unusual to be able to take visual bearings on terrestrial landmarks but when they are available, navigators take advantage of them. I'll bet that any WW2 Air Force navigator that flew over the western Pacific would tell you the same thing, that he used every opportunity that presented itself...

I don't doubt that - from beginning to end of flight ("DON'T STOP NAVIGATING"), but I also wonder how often they actually sent birds zig-zagging like that to do it.  Given a healthy octant (maybe it wasn't by then - maybe FN dropped it and wanted to check it again, who knows) FN shouldn't have had so much concern at that point.

On the whole, I can see your points and while it works fine and there are some advantages, I just not compelled to buy that they did this.  Anyway, if they did, I'm still not clear on what effect it would have had on the outcome - except to somewhat improve their chances by improving the tolerances for DR and giving one additional assurance that the octant was up to snuff.

I've learned alot from studying this.  You remind me of a couple of great instructors I had over the years - stand-out guys who encouraged students to learn past the basics and put some good extras in their tool box - great stuff and it makes a difference, especially when things take a turn down sometime.

I just saw a taped interview of the surviving guys who shot down Yamamoto while I was visiting the Air Museum in Seattle.  That flight of AAF P-38's DR'd hundreds of mile out over open water - dog-leg pattern to first avoid lanes and land, and then back in to intercept - with nothing more than a watch, strip charts and a borrowed nautical compass from the local Navy guys (I wondered what was wrong with the whiskeys - but I gather that borrowed compass was probably stabilized, etc.).  Of course that's less distance, the P38's faster, and those guys were experienced night fighters, but it's still a heck of a feat done by pure DR over a lot of open water.

LTM -

[Gary LaPook]

Maybe you didn't understand how flying over Nauru would allow checking his navigation and the accuracy of his octant.  All celsetial navigation involves picking a position on the ground and calculating what altitude of a celestial body would be measured from that spot...

There are additional reasons to fly over Nauru. One is to determine the winds at their cruising altitude. You were trained to compute a wind correction angle using a wind vector diagram on your E6-B but few instructors teach the next computation, calculating the winds encountered in flight by using the same vector diagram... Using the Ontario for the visual checkpoint would not allow the same accuracy because of the uncertainty in the position of the ship.

Yet another reason why flying over Nauru was desirable is that it would provide an accurate starting point for the dead reckoning to Howland. The accuracy of dead reckoning can never be any better than the accuracy of the starting position...using the Ontario as a starting position has even more uncertainty than a celestial fix because Noonan could not know, for certain, the actual position of the ship. And, he had to be aware of the possibility of another ship being in the area...

Navigators are expected to use all information available to them for safely conducting the flight. In trans-oceanic navigation it is unusual to be able to take visual bearings on terrestrial landmarks but when they are available, navigators take advantage of them. I'll bet that any WW2 Air Force navigator that flew over the western Pacific would tell you the same thing, that he used every opportunity that presented itself to take a visual observation of islands to confirm and to improve his navigation.

So, compared to the slight cost of less than an additional 10 NM, it was well worth going over Nauru.

gl

I see your points - and these things are logical. 

But -

...flying over Nauru would allow checking his navigation and the accuracy of his octant.  All celsetial navigation involves picking a position on the ground and calculating what altitude of a celestial body would be measured from that spot..."

Shouldn't his octant have been OK after such a check at say, Lae, and perhaps again as the flight passed the Nikumanu Islands?

Sure, and I would bet money that Noonan did the same kind of sextant check while on the ground at Lae. But it would be a good thing to make another check as late as possible, just before beginning the last, most critical leg, since proceeding along the last leg, past the PNR, was an irrevocable decision. In addition to providing a check of the octant this also provided a check on the chronometer and the tables. He couldn't do the same at Nukumanu because the sun was too low to be observed with any level of accuracy.

One [good reason] is to determine the winds at their cruising altitude.

Wouldn't that have been possible as the flight passed the Nikumanu Islands?

Yes, and Noonan did. That is where they radioed to Lae that the wind was 23 knots most likely computed as a "wind between fixes" starting from a visual fix over Choiseul and ending with a visual fix at Nukumanu. This is why they, again, made a slight deviation, only five minutes of extra flying time, from the straight line from Choiseul to Nauru in order to get an accurate visual fix on Nukumanu to get an accurate wind.

...another reason why flying over Nauru was desirable is that it would provide an accurate starting point for the dead reckoning to Howland...

I realize that Nauru would put them closer and therefore could reduce the potential error - good point, but again, wouldn't passage past Nikumanu have given confidence and a reliable 'starting' point for DR?

It's 1492 NM ("Columbus sailed the ocean blue") from Nukumanu to Howland producing a DR uncertainty of 149 NM while it is only 993 NM from Nauru producing a DR uncertainty of only 99 NM, 50 NM less than from Nukumanu.  But think about how this plays out in the end game, as Noonan was paid to do. When they intercepted the sunline LOP, the 99 NM uncertainty from Nauru would require searching along the LOP for twice the DR uncertainty, for 198 NM total to be certain that they had made a complete search. Increasing the DR uncertainty by 50 NM would then require searching along the LOP for 298 NM, an extra 100 NM. And that is if they used the offset, landfall, procedure (which I believe they did.) But, if they did the "straight in and then search both ways along the LOP" procedure, as Ric and others believe, then the extra uncertainty would add an additional 150 NM making it 447 NM total flying along the LOP instead of 297 NM if they had come from Nauru. Well worth the extra 5 minutes flying time to go over Nauru.

Navigators are expected to use all information available to them for safely conducting the flight. In trans-oceanic navigation it is unusual to be able to take visual bearings on terrestrial landmarks but when they are available, navigators take advantage of them. I'll bet that any WW2 Air Force navigator that flew over the western Pacific would tell you the same thing, that he used every opportunity that presented itself...

I don't doubt that - from beginning to end of flight ("DON'T STOP NAVIGATING"), but I also wonder how often they actually sent birds zig-zagging like that to do it.  Given a healthy octant (maybe it wasn't by then - maybe FN dropped it and wanted to check it again, who knows) FN shouldn't have had so much concern at that point.

On the whole, I can see your points and while it works fine and there are some advantages, I just not compelled to buy that they did this.  Anyway, if they did, I'm still not clear on what effect it would have had on the outcome - except to somewhat improve their chances by improving the tolerances for DR and giving one additional assurance that the octant was up to snuff.

I've learned alot from studying this.  You remind me of a couple of great instructors I had over the years - stand-out guys who encouraged students to learn past the basics and put some good extras in their tool box - great stuff and it makes a difference, especially when things take a turn down sometime.

I just saw a taped interview of the surviving guys who shot down Yamamoto while I was visiting the Air Museum in Seattle.  That flight of AAF P-38's DR'd hundreds of mile out over open water - dog-leg pattern to first avoid lanes and land, and then back in to intercept - with nothing more than a watch, strip charts and a borrowed nautical compass from the local Navy guys (I wondered what was wrong with the whiskeys - but I gather that borrowed compass was probably stabilized, etc.).  Of course that's less distance, the P38's faster, and those guys were experienced night fighters, but it's still a heck of a feat done by pure DR over a lot of open water.

LTM -

It was about a 400 NM flight so they could expect to hit their coast-in point within 40 NM, and most likely less. They were aiming for a large island so had the opportunity to correct their flight path when they made landfall. I suspect the compasses borrowed from the Navy were the aperiodic type that I have mentioned before.

[JNev]

...Contrary to the  "don't know how far north or south they are" argument, Noonan certainly did, just by dead reckoning... ...the maximum expected D.R. error is 110 NM (128 SM) so they would not have proceeded more than 110 NM south from the D.R. position of Howland before ...going back to the north, searching for Howland.

That explains one constraint nicely, thanks.  It also underscores why Nauru is important in your thinking - it could narrow the error by closing the DR distance. 

A N-S determination by celestial (sun) would be limited to a shot while at meridian - too late to do the flight much good, is that correct?  (See "American Air Navigator", Mattingly (1944), page 158 - your site.)  That emphasizes need for DR as to course.  <Ah - I came back to correct something I deleted accidentally and thought of 'shoot the moon' further down - never mind this point.>

...it is much more likely that they got a fix around 1623 Z, or even later, making the maximum D.R. error only 46 NM, see Landfall procedure navigation to Howland Island.

1623Z or later - I'll buy that.  Per your article "we can assume Noonan was busy right up until the time of civil twight..." - I agree that fits "NEVER QUIT NAVIGATING".  It also gave FN about an hour and 15 or so minutes before civil twighlight (where the flight should have been at that time).

AE's call at 1623Z reportedly included "partly cloudy", so a shot may have been delayed (may fit "or later"); at 1742Z AE reports "200 miles out" - just when civil twightlight should have arrived at her position - FN may have just gotten a last shot in to establish that distance.  If he had multiple bodies in view he also could have established other lines, yes?  That would help fix N-S at least at the point of having such shots.

... Noonan knew how far they had flown since the last fix and would have allowed the appropriate offset for the intercept point on the sunline LOP.  So, even if it was just dead reckoning all the way from the Ontario, then Noonan would have aimed 110 NM, at least, to the north-northwest of Howland which would ensure that they did not end up south of Howland at the point of intercept.[/i]  They would then fly 220 NM south-southeast along the LOP looking for Howland so, worst case, if they missed the island and if they had been at the maximum D.R. error to the right point of interception, they would still not proceed more than 110 NM further to the south-southeast before turning back to the north to execute a search pattern.

Good plan, but why so certain NNW and not bias to SSE intercept?  A SSE intercept would be a bit further (LOP tangent), but it would put the error toward back-up landfall - Baker for one, also Phoenix group.

I realize we're probably going to disagree, but your logic on this point runs counter to what Friedell (CO - Colorado) was being advised of, and believed was likely -

"...first despatch cast definite doubt as to the location as being 281 miles north of Howland... the region to the north of Howland... was entirely water. ...other despatch referred to the opinion of the technical aides connected with the flight, that the plane would be found in the original line, which would indicate a position through Howland Island and the Phoenix Group... These reports bore out the original assumption of the Commanding Officer, which was based on all information then available, that the logical quadrant for the position of the plane was the southeast quadrant...

...Considering the question as to what Mr. Noonan did do, it must be considered which way he would steer on the line. To the northwest of Howland was wide stretches of ocean, to the southeast were spots of land. To a seaman in low visibility the thing to do when in doubt of own position would be to head for the open sea. The land would be the place to get away from. To the Air Navigator with position in doubt and flying a land plane it is apparent that the thing to do would be to steer down the line towards the most probable land..."

So, at least one case was made at the time for SSE of Howland; if FN was doing as Friedell and others suspected, a SSE bias makes more sense to support landfall, failing Howland for any reason.

Consider a 're-aim' -

...110 NM, at least, to the north-northwest SSE of Howland [/b] which would ensure that they did not end up south north of Howland at the point of intercept.[/i]  They would then fly 220 NM south-southeast NNE along the LOP looking for Howland so, worst case, if they missed the island and if they had been at the maximum D.R. error to the right point of interception, they would still not proceed more than 110 NM further to the south-southeast NNE before turning back to the north to execute a search pattern.

 
- Strikethroughs and color indicate my changes to your quote.

By the way, I share Friedell's notion in part because as you've said earlier today "Navigators are expected to use all information available to them for safely conducting the flight", and because other seasoned people, such as Friedell, understood it.  The Phoenix group would have been a known thing, as would Baker, to a well prepared navigator doing as you've suggested he should.

The problem is  and what AE did say ('on line' - never 'in box'), I am moved toward "LOP" as the 'search pattern' and away from box followed by LOP.  On your site, "American Air Navigator", Mattingly - 1944 - pages 157 - 158 discuss both the box search and LOP approaches to landfall.  (Single Line of Position Landfall Position) - the excerpt you provide there starts out by covering a box search procedure (example: Wake Island), and then moves on to a LOP landfall procedure (which happens to use a Canton Island example, of all things...) - the latter exercise fitting at least the "on the line" LOP call.

[/b] But what if... a much larger error in the DR than expected, wouldn't they have ended up much further south? ...that brings in Baker, 38 NM south of Howland. In order to miss seeing both Howland and Baker, and with 20 NM visibility, they would have had to have been an additional 58 NM off to the right of the DR course  in order to pass so far south of Baker so as to not be able to see it.

Pretty strong assumptions about being able to spot those two islands.  Whatever you believe, many see Howland as a challenge - and Baker was probably no better.  Maybe could be seen for 20 miles, maybe not.  I don't doubt that conditions generally weren't bad - just don't believe that it would be so easy to spot those two places because of their physical character, for one thing.

This would be a total DR error of 168 NM, 15% of the distance flown from Ontario and 26% of the distance flown from a 1623 Z fix. It is highly unlikely to have such large DR errors. Based on the statistics of navigation, (appendix Q in the 1977 edition of the American Practical Navigator) there is only one chance in 370 of being 15% off course and only one chance in ten-million of being 26% off course! And, as is likely, if Noonan added an additional safety margin to his offset then missing both Howland and Baker is an even more remote possibility.

Agree, but miss somehow he did.

Now, the second point. If the DR accuracy doesn't convince you then let's shoot the moon. I have pointed out many times that the moon was positioned to provide an LOP that would tell Noonan whether he was north or south of Howland and so would also prevent flying down to Gardner.

Well, it wouldn't "prevent flying down to Gardner", Gary - but it should have allowed him to find Howland - which he didn't.  He could still fly "down to Gardner" after failing to find Howland, and apparently Baker.

Looking at 1912 Z, the height of the moon was 74° 26' at Howland and its azimuth was 328° which produced an LOP running 058° -238° T. (We know that Noonan could take observations at least as high as 75° since he did so on the leg to Hawaii.) Using this LOP, Noonan would have known how far he was north or south along the 157° -337° sun line LOP. I have attached a chart showing a fix using the sun and the moon at 1912 Z. (I am not saying that they were at this fix position, this is just an example of a fix that Noonan could have obtained at 1912 Z.) So, looking at the moon LOP running from the lower left to the upper right, you can see that Noonan could have determined how far they were south of Howland and so would have let then know that they had to turn around to go back to the north to search for Howland.

Excellent.  So any of at least 3 things may have happened:
- This shot was not available for some reason we can't understand
- Something failed in the process of applying the fix, or
- The navigation worked very much as you believe and the flight came tragically close, but through a combination of error (celestial ain't perfect - and DR's less so) and visibility (due to distance and / or local conditions), neither Howland nor Baker were spotted. 

The first may be a tough-luck possibility,
The second is unlikely IMHO, and
You are very convincing as to the possibility of the third, excepting that we'll probably disagree on ability to see the islands of Howland and Baker.

The yellow moon LOP was calculated from an observation of the moon of 73° 22' placing the LOP, and the observer (Noonan)  64 NM south-southeast of Howland. The white sunline LOP was calculated from an observation of 17° 13' placing this LOP and Noonan 109 NM west-southwest of Howland. From the plotted example fix it is 153 NM on a course of 032° T to Howland. The weather conditions south of Howland were conducive to celestial observation of both the sun and the moon.

So they may have come very close to Howland, and Baker for that matter - but seem to have spotted neither.  What's to prevent the flight from bugging out for Gardner down the line of position after that 1913Z call?  It appears to me that contrary to where we were on the MC analysis a couple of days ago, you've gotten this flight right back on top of Howland. 

How far west of the LOP is Gardner (Niku) - about 15 miles?  Doesn't that leave Gardner within scope on the other end of a 300 mile DR course?  I don't see the problem with that. 

I do see a problem with spotting the cow patty islands of Howland and Baker; far less so with bright, blue lagooned / wide fringing reefed Gardner with her 90 foot Buka forests on the necklace.  Just IMHO - YMMV.

EXCELLENT exercise too Gary - I'm fascinated by all you've shared.

LTM -

[Harry Howe, Jr.]

Simply an opinion
FN was probably  inebriated after 2 nights/days partying with the guys at Lae.  He prolly wasn't sober when he got on the plane for takeoff.  He probably was asleep (or passed out) during the early stages of the flight (perhaps longer) and AE was on her own to fly, radio, navigate, etc.

AE's telegram citing "personnel unfitness" was her way of telling George what was happening and the Brine's letter more than hints at it.

He (FN) prolly had more than one bottle of his favorite liquid refreshmnt along in his kit and prolly took a nip or two or more along the way.  Just an opinion.

[JNev]

Sure, and I would bet money that Noonan did the same kind of sextant check while on the ground at Lae. But it would be a good thing to make another check as late as possible, just before beginning the last, most critical leg, since proceeding along the last leg, past the PNR, was an irrevocable decision. In addition to providing a check of the octant this also provided a check on the chronometer and the tables. He couldn't do the same at Nukumanu because the sun was too low to be observed with any level of accuracy.

One [good reason] is to determine the winds at their cruising altitude.

Wouldn't that have been possible as the flight passed the Nikumanu Islands?

Yes, and Noonan did. That is where they radioed to Lae that the wind was 23 knots most likely computed as a "wind between fixes" starting from a visual fix over Choiseul and ending with a visual fix at Nukumanu. This is why they, again, made a slight deviation, only five minutes of extra flying time, from the straight line from Choiseul to Nauru in order to get an accurate visual fix on Nukumanu to get an accurate wind.

...another reason why flying over Nauru was desirable is that it would provide an accurate starting point for the dead reckoning to Howland...

I realize that Nauru would put them closer and therefore could reduce the potential error - good point, but again, wouldn't passage past Nikumanu have given confidence and a reliable 'starting' point for DR?

It's 1492 NM ("Columbus sailed the ocean blue") from Nukumanu to Howland producing a DR uncertainty of 149 NM while it is only 993 NM from Nauru producing a DR uncertainty of only 99 NM, 50 NM less than from Nukumanu.  But think about how this plays out in the end game, ...for 198 NM total to be certain that they had made a complete search. Increasing the DR uncertainty by 50 NM would then require searching along the LOP for 298 NM, an extra 100 NM. And that is if they used the offset, landfall, procedure (which I believe they did.) But, if they did the "straight in and then search both ways along the LOP" procedure, as Ric and others believe, then the extra uncertainty would add an additional 150 NM making it 447 NM total flying along the LOP instead of 297 NM if they had come from Nauru. Well worth the extra 5 minutes flying time to go over Nauru.

P-38's / Yamamoto...

It was about a 400 NM flight so they could expect to hit their coast-in point within 40 NM, and most likely less. They were aiming for a large island so had the opportunity to correct their flight path when they made landfall. I suspect the compasses borrowed from the Navy were the aperiodic type that I have mentioned before.

Very good points, Gary - I can see the advantages of a Nauru deviation now, thanks for taking time to go into all that so well.  That should have put the flight just about on top of Howland and therefore leaves the mystery that much deeper somehow.  I get the feeling NR16020 was skimming the horizon just out of reach before moving on to eternity.  Well, surely AE and FN did - I still hope NR16020 is anchored to this earth and gets found yet. 

I'd bet you are right about the P-38's - meant to go back and look at what you wrote about the aperiodic compass. 

Thanks, Gary.

LTM -