Print

[JNev]

...Logic.  All of the alternative islands are to the southeast.  Use your remaining fuel heading in the direction that could bring you to an island. 

I second that.  Weems wrote of the underlying technique in 1938.

Also, she said, "Running on line north and south" not "south and north."

Ric

I agree that the statement in that order is intriguing, but for me it is not strong enough alone to indicate which direction was final.  A statement like 'running on the line north, then will run south' would have meant more, but we don't have it, so we are left with our best interpretation.

To me, the strength of it lies with the statement - which affirms the running of the line as it makes sense and we believe it to have been, PLUS the full context of couching her statement within the underlying navigational logic.  While it has been argued that the Phoenix groups is more seive than catcher's met due to islands being so scattered, it still represents far more chance of landfall than would have proceeding to NNW of Howland; if one takes the chance in a landplane, one would likely err to the SSE on the final leg, IMHO, based on the understandings of the day as best we can know them.

I respect the other ideas - nothing is provable so far, but to me a final leg SSE along the LOP makes the most sense given what was last said and by what we can observe of the navigational and range picture.

[Ric Gillespie]

Actually, the best indication that they turned SSE is the abundance of hard evidence suggesting that they ended up on Gardner.  Had they turned SSE first and then NNW it's hard to see how they could have reached Gardner.

[John Ousterhout]

Every time I read a rehash of the communications I get a different impression.  That's one of the reasons I enjoy these discussions.  I'm afraid I'm a bit obsessed imaging what went on in the cockpit.
I can imagine this: Amelia is flying along, receiving nothing but static on the radio.  Fred tells her they've arrived on the line.   She transmits "We must be on you but cannot see you".
She continues to hear nothing but static.  She fiddles with the receiver some more, searching up and down frequency.  She switches to the loop antenna.
She fiddles some more and finally gets clear reception of code, understanding that it must be from the Itasca.  She transmits "We received you but cannot get a minimum...".
Her initial assumption upon finally hearing Itasca is that the reason she can finally hear Itasca is because she thinks they're approaching Howland, not because of the antenna change.  They continue flying in the same direction but hear nothing more (and have switched to a different antenna).
They fly some more, continuing along "the line".
She concludes they've flown out of range, therefore away from Howland, so Howland must be back the other way along the line.  They/she stops believing celestial navigation and starts believing her misunderstanding of the radio signals.  She doesn't know about the broken antenna, nor the donut hole.
They turn and fly back the way they came, continuing to transmit and listen (on the wrong antenna and/or frequency), expecting to pick up the Itasca's broadcasts again, sure that if they continue "on the line" they'll eventually get back within radio range and that will solve their problems.  That's why they fly the wrong way, even though Fred's sun and moon sights say they should go the other way.
How many aircraft accidents have occurred because the pilot didn't believe the instruments that gave truthful information?
Comments?

[JNev]

Every time I read a rehash of the communications I get a different impression.  That's one of the reasons I enjoy these discussions.  I'm afraid I'm a bit obsessed imaging what went on in the cockpit.
I can imagine this: Amelia is flying along, receiving nothing but static on the radio.  Fred tells her they've arrived on the line.   She transmits "We must be on you but cannot see you".
She continues to hear nothing but static.  She fiddles with the receiver some more, searching up and down frequency.  She switches to the loop antenna.
She fiddles some more and finally gets clear reception of code, understanding that it must be from the Itasca.  She transmits "We received you but cannot get a minimum...".
Her initial assumption upon finally hearing Itasca is that the reason she can finally hear Itasca is because she thinks they're approaching Howland, not because of the antenna change.  They continue flying in the same direction but hear nothing more (and have switched to a different antenna).
They fly some more, continuing along "the line".
She concludes they've flown out of range, therefore away from Howland, so Howland must be back the other way along the line.  They/she stops believing celestial navigation and starts believing her misunderstanding of the radio signals.  She doesn't know about the broken antenna, nor the donut hole.They turn and fly back the way they came, continuing to transmit and listen (on the wrong antenna and/or frequency), expecting to pick up the Itasca's broadcasts again, sure that if they continue "on the line" they'll eventually get back within radio range and that will solve their problems.  That's why they fly the wrong way, even though Fred's sun and moon sights say they should go the other way.How many aircraft accidents have occurred because the pilot didn't believe the instruments that gave truthful information?
Comments?

"They/she stops believing celestial navigation and starts believing her misunderstanding of the radio signals.  She doesn't know about the broken antenna, nor the donut hole." -

"That's why they fly the wrong way..." -

That's galvanizing somehow, John.  I've long tried to digest Gary LaPook's navigational points and suddenly a few things are jumping out at me -

My understanding is that the moon was available, and that an observation of the moon by Noonan should have instantly told him that the flight was south of Howland, and how far south.  My thought on that is either Fred or the moon is missing here.

Back up a bit in the flight and consider the long-distance DR prospect.  As has been said here, more or less, the accepted level of uncertainty for a position found by dead reckoning is about 10% of the distance flown - since the last fix.

- Dead reckoning all the way from Lae involves around 2556 statute miles +/-, so a 10% error would be in that case around 230 nautical miles or 260 statute miles.  That would be the expected outcome if the flight had had to slog all the way from Lae to Howland in clouds, without benefit of fixes along the way. 

But the flight didn't having to face that, so far as can be realized - the Ontario and other visual fixes should have narrowed that problem -

- The flight had a fix at 0718 Z near Nikumanu Island - still 1700 SM from there to Howland; expected uncertainty from there to Howland would be about 170 SM;
- Then we believe they may have seen the Ontario at 1030 Z - about 1270 SM from Howland; uncertainty by the time of arrival at the 1912 Z call would then have been only about 127 SM;
- Then the flight passed Nauru at about 1130 Z - about 1143 SM remaining to Howland; uncertainty drops to 115 SM by the time they would reach the call point at 1912 Z;
- We believe they flew over Tabituea - only 613 SM remaining to Howland; that drops the uncertainty by dead reckoning to about 61 SM.

Of course we have to consider some additional odds - the 10% assumption is not 100% reliable, so to speak.  I'm no whiz, but do get the basics on some normal curve stuff - dead reckoning involves uncertainty, already stated at 10%.  Beyond that, we have to realize that our 'band of uncertainty' (the 10% error) should contain around 95% of the possible
actual positions of the aircraft; there remains about a 5% chance that the flight would be outside that 10%
band - or a 5% chance of being further off course than the 10% illustrations made above.

In standard deviation terms, 95% equals 2 standard deviations meaning that one standard deviation was only half of the band of uncertainty. As you exceed this distance the probability that you are further away decreases very quickly.  I'll avoid embarrassing myself with the actual math - but suffice it to say that the chances of a gross error diminish rapidly as a flight departs the normal error range. 

Consider then - from Lae to Howland in the 'blind' would equal about a 255 statue mile error possibility.  To break that down statistically and within the standard deviation 'wedge', 68% of the "95 percentile" time you will wind up within half of the uncertainty band - or be no more than about 128 SM off; 32% of the "95 percentile" time you may wind up somewhere between a half and 1 full standard deviation (the 'outer one-half of the error band').  So, the uncertainty at 1912 Z can reasonably be seen as 255 SM - which is 2 standard deviations (95% certainty total), so one standard deviation = 128 SM, a 68% possibility.  The chances of an error approaching the 255 SM / 230 NM assumption are on the diminishing end of that 'outer band' of 32%... the chances of being further off than that also diminish and much more rapidly - somewhere from 5% and diminishing, the further 'out' you go.

But we do not have a reason to believe that the flight went blind all the way from Lae to Howland vicinity - we have reports as outlined above.  Forget the closer Tabituea (I am thinking that was a ground report of a flyover) - only 613 SM remaining to Howland, which would drop the uncertainty by dead reckoning to about 61 SM.  Consider only the Earhart reported fix at 0718 Z near Nikumanu Island - 1700 SM remaining to Howland: the expected uncertainty would be about 170 SM.  That known point should then drop the uncertainty to 170 SM by the time they would reach the call point at 1912 Z. 

Assuming a decent fix at Nikumanu Island, that means a 68% chance of half that 170 SM error - or about 85 SM, and up to a 32% chance of the full 170 SM error.  Of course we can't really count on being within 2 standard deviations 100% of the time (68% + 32% = 100% - NOT, but IS 100% of 95%...), so we have to accept some chance of arriving outside that range of error (5%, as said above).  So, an assertion of 255 SM, or 230 NM 'off course' is about 1.5 times the generally accepted uncertainty, given the Nikumanu Island fix, or about 3 standard deviations.  If I'm following the tabular stuff in my old book correctly, then that means the odds of arriving 'on the line' 230 NM south of Howland are at about one in 370.  It can happen - but betting that way can lose a lot of money for you, too.

All numbers, all probabilities, all based on what we know of Dead Reckoning.  I believe the Nauru passage is reliable as a report from Earhart, but if I am wrong I will happily stand corrected.  There are other fixes I believe, so one need only do the basic math from those points and YMMV, of course - but odds is odds.

I have to say that I have underappreciated this point in the past because the LOP / down the line to SSE seemed so obvious to me.  Well, it did to the navy as well, no sin.  The problem with it is, if you look at 'the line' and how Howland, Baker and Gardner 'line up', one MUST fall far enough south to make the northerly turn and go long enough to break a sweat about fuel, and turn back, having NOT spotted even Baker island. 

Baker is smaller and uglier than Howland - and yes, we've spoken of cloud shadows - more 'variables' lurk in there, so I cannot speak in absolutes here.  But the point is, in all likelihood, to hit the LOP far enough south for that to happen with any substantial likelihood puts the flight way outside of the expected DR error - way out on a skinny limb in terms of 'odds'.  I must consider that there may be another and stronger answer for how such a thing would happen.

Now comes John's infernal point here, which drove me back into all this stuff.  Turns out I have to say that I have also overlooked 'the moon', which Gary kept howling at (sorry Gary, if you read this ) and now I'll say I may have been a bit too dismissive.  The moon was there for the reading - if it could be seen. 

So which was missing, Fred?  Or the moon?  And of course, there could be other variables... a dropped / broken instrument (octant and / or sextant 'preventer', if aboard - argued both ways by things Fred had written in the past and other records,etc.) or some other reason for failing to get a reliable reading.  Not sure where those fall out in terms of odds, but given reasonable DR navigation and the moon, Fred should have avoided a major southerly error.

Now I start thinking of Earhart's mention of clouds - and clouds to the north and west of Howland.  Clouds are one reason Fred might not see the moon.  Despite the evidence we have at Niku, which to be objective and rigorous in my analysis I have to argue is circumstantial and that it is not an 'abundance of hard evidence', I have to keep an open mind.  In truth, I think, despite much hard and good work and many fine things before us to consider, we are still having to live within the realm of possibilities of a landing at Gardner (now Niku) - perhaps even a truly tough set of probabilities. 

What of the other points - the pre-war dural skin, the window shard?  Those are two of my favorites and definitely of great interest as Electra-sourced possibilities.  What of the castaway and items we associate with a female?  Compelling, but how much so depends on how accepting one is - they remain circumstantial.  The Bevington Object is compelling - I certainly believe it to be an 'object', and I see things there that do suggest 'gear' - but I still have to accept that I'm 'living in the odds'.  Etc.

Radio evidence - again, compelling to me - but still out there in the odds.  The tabular odds show some interesting patterns to me - but they are still pretty tough odds.  I cannot show where Dana Rudolph and Betty Klenck could have most definitely been mistaken by commercial programming of some sort, but I'm not sure we can entirely rule it out.  Compelling, interesting - but still circumstantial.

The Pan Am DF receptions - this has been a very interesting and compelling point for me.  Those were professional radio guys listening in, with airline state-of-art equipment of the day, I'm sure.  There should not have been a lot of sources on that frequency bouncing around in that area - Earhart seems a good explanation.  Maybe so - and if so, had to be on land.  Mantz thought she would be, so did the navy. 

Well, the navy looked - nothing.  Eventually the navy did what most organizations do when they have to close an issue - outliers like unvalidated radio traffic gets explained as best can be done - rightly or wrongly.

The problem is simply 'odds'.  We're still struggling with long odds, IMHO.  I followed the 'markers' at Niku for a long time - still will - but am wondering if I should also pay more attention to others. 

The navigation case just is not so simple.  If Fred was so good and all that methodology was so great, then why didn't they land at Howland?

I have to realize that they still could have tragically crashed at sea just north or west, short of Howland, despite the odds I've thought of for a long time.  All kinds of cases can be constructed for hypothesis.  I have to accept that 'clouds' may have played a larger role than I've previously thought of - they may have obscured the stars and even the moon; they may have blotted out a faintly distant Howland in the herd of shadows from them. 

Maybe there were clouds down toward Niku too - and we just don't know it.  We do know there were clouds to the north and west of Howland, per the Itasca.  We also have the Itasca's skipper's first reaction - to search in that direction: one lowly cutter of the day, out looking on a broad expanse of ocean for a downed plane.  It may have gone down out there - may have floated for days, unseen - or gone straight to the bottom in a crumpled mess.

Or it may have gone 'down the line' to Gardner.  But I have to accept 'the odds', still.

Something to think about.  It's been a long, long search out there at Niku - no doubt more so for Ric and a few others than for the rest of us.  There have been many tantalizing things found, but unless one is willing to take the aggregate circumstantial evidence as conclusive, we still lack the holy grail.  The only way to know is to find it.  I guess where to look depends on how one is willing to bet, IMHO.

[John B. Shattuck]

Jeff,

Enjoyed your analysis, but I have to wonder if statistical analysis is a valid method for analysis here.  Not to sharpshoot, and I look forward to your thoughts and response; but we do not have a population of 30 or more airplanes flying from Lae to Howland... we have one.  Whatever the odds of it arriving, it did not...it did arrive somewhere, and the physical evidence; while maybe not conclusive, is consistent with a landing at Niku. 

I understand you were looking at the miles potentially off by the natural error in dead reckoning; and again we have a population of one.  No way of knowing how many standard deviations this population of one may have fallen in the population of error probabilities; and IMO the point is moot when one considers that we do not know how far they may have been dead reckoning...if at all (okay, certainly they were between celestial sightings but you know what I mean ).

IMO, sometimes we all get caught up in what would have, should have, maybe did, etc.  But in the end, we have to follow the evidence to a conclusion.  The evidence seems to be on Niku, whatever the statistical probability of ending up there.

Respectfully,

JB

[JNev]

Jeff,

Enjoyed your analysis, but I have to wonder if statistical analysis is a valid method for analysis here.  Not to sharpshoot, and I look forward to your thoughts and response; but we do not have a population of 30 or more airplanes flying from Lae to Howland... we have one.  Whatever the odds of it arriving, it did not...it did arrive somewhere, and the physical evidence; while maybe not conclusive, is consistent with a landing at Niku. 

I understand you were looking at the miles potentially off by the natural error in dead reckoning; and again we have a population of one.  No way of knowing how many standard deviations this population of one may have fallen in the population of error probabilities; and IMO the point is moot when one considers that we do not know how far they may have been dead reckoning...if at all (okay, certainly they were between celestial sightings but you know what I mean ).

IMO, sometimes we all get caught up in what would have, should have, maybe did, etc.  But in the end, we have to follow the evidence to a conclusion.  The evidence seems to be on Niku, whatever the statistical probability of ending up there.

Respectfully,

JB

Thanks John, you are very kind.

Statisitcal probabilities absolutely apply, John - and to that point, I've not said 'would have' or 'should have', etc. - but have merely pointed out some rational numerical realities. 

The prospect is also very much 'one airplane, one flight, one likely outcome' and probabilities of same, not a phantom fleet or repeated efforts.  Who knows within a target ring a dart will land?

Ric has assigned, for hypothesis' sake, a probability of 'one' to the airplane arriving 'down the line' by 230 NM, if I follow him correctly.  In simple terms, that means a 1 in 370 set of odds for such an arrival on a given flight.  This does assume the established fix of 0718 Z near Nikumanu Island - still 1700 SM from there to Howland as a reliable fix (I believe the record supports that), and I picked one 'further out from Howland' to inject some extra error margin in favor of Ric's hypothetical assumption. 

That means, in simple dead reckoning terms, an expected possible error from there to Howland Island of about 170 SM.  The point of the statistics is that we have a 95% probability of arrival within that range of error.  We have a 5% chance of arriving even further away from Howland, and that likelihood diminishes with each mile further away considered.

That further means that a hypothetical assertion of 230 NM / 255 SM 'off course' is around 1.5 times the generally accepted uncertainty of "10%" from the last fix (Nikumanu), or about 3 standard deviations.  In turn, that means in simple terms that our one airplane, on this one flight has odds of about 1 in 370 of arriving 'on the line' 230 NM south of Howland.

BTW, I did not use celestial 'fixes' but rather a land fix that was reported - there's a difference (more accurate, I believe).

Can that happen?  Of course it can happen - I never said it 'could not happen' - I've merely illustrated the 'odds' given standard DR navigation from a fixed point that was reported as observed during the flight. 

So if the flight arrived as Ric has hypothesized, then it did so against the more likely outcome ('against the odds' so to speak) of 'not being more than about 170 miles away' from Howland.

Did it arrive somewhere?  Yes, I heartily agree and believe we all can - sans aliens, it DID arrive somewhere in the Pacific area - absolutely 100% certain of that, an 'event' of 'one', for sure. 

I have also been one to point out many times that TIGHAR has found and reported on far more (in fact, so far as I can tell, "100% more") tangible 'stuff' supporting a reef landing, subsequent cries for help, a near-miss aerial search and a castaway that could have been Earhart on Gardner (Niku).  Consider Ric's point on this - similar to my own:

...best indication that they turned SSE is the abundance of hard evidence suggesting that they ended up on Gardner...

Ric doth not lie - he speaketh truth.  I merely parse to make the point that we have two operative conditions in that phrase:

The first fills me with optimism -

- Abundance of hard evidence - yes, we have 'things in hand' that are tangible, and taken in full 'context' as I have often advocated myself, they seem to tell a 'story' - one of Earhart coming to her end among the Strawberry and Coconut Crabs of Gardner.

But we're still stuck with the second operative condition in that phrase - and it amounts to 'statistics' in its own way -

- Suggesting - this is brutal honesty; this is why I've never really had a problem with TIGHAR's approach to establishing and testing a hypothesis about a reef landing at Niku, and hat's off - it reminds me to remain rigorous in my analysis. 

It also bequeaths an unfortunate reatlity: I am stuck with a suggested outcome, i.e. an outcome based on a set of probabilities when one thinks of it.

Is it a high probability?  Ric suggests it is a high probability because of the physical evidence - which despite all I wish for, remains toughly 'circumstantial' for now in my view.  YMMV, and I argue not - I merely point out realities as I can understand them.

The Niku hypothesis also seems more clearly now to hinge at least to some degree (damn statistics again) on an assumption of the flight arriving 'on the line' some 230 NM SSE of Howland - at odds of 1 in 370 from my view as I have now carefully considered the navigational case. 

Starting with that set of odds, now I have to consider the risk that the 'abundance of hard evidence' could actually be of origins other than Earhart, although I consider that it does strongly suggest an Earhart presence.  Can it all have happened?  Of course it could have - despite the odds.  It is a big world, strange things happen.  Can it be proven by these things we hold that she was there?  I've yearned for it - still do, heart aches for Ric to lay hands on something that will prove it - but so far I fear we are not there.

Is is likely to have happened?  The odds are terribly inconvenient - from the navigational realities right down to the radio traffic probabilities, etc. 

My point really is - until somebody emerges with a chunk of positively identifiable material from Earhart's flight that can only have arrived in a given place, within reasonable bounds of 'odds', we don't know for certain.  And, it is a large world full of possibilities - while Niku is one of them, so are other outcomes. 

---

The corollary to my point is 'choose your poison' - if you believe in Niku that is no sin in my book - it is a wonderful hypothesis and has been thoroughly examined for something approaching 30 years.  In parallel to that, be careful about condemning other ideas - the clouds that day to the NW of Howland may have played the stinker - Fred may well have been denied a moonshot by those clouds, and the flight could easily have come within miles of Howland but the fliers may have missed smoke, island shape and all among the morning ocean and cloud shadows.  They may have, for all we presntly know for certain, have died in a crash a few miles NNW of Howland and flying straight at her 'on the line' with fule exhausted and muted by a day frequency, despite our estimates of endurance.

Conclusively, the best searches to date have yielded something we can bank on: so far, we know exactly where the airplane ain't.

Statistically speaking in terms of the region, the Pacific is a large place that can quickly swallow large items without a trace.  In my view, to be realistic we have to at least give some credence to the 'numbers' or odds.  I dislike the inconvenience of them as much as anyone.

So, choose your poison and don't sweat it - but accept the odds when you do.

[JNev]

So after all this time I've only just found out that there was a moon available!

A moon would slow Fred to get a fix on their position so it begs the question: If they missed Howland what went wrong?

My understanding of a single fix on the moon is that Fred could have determined how far north or south he was - time vs. altitude of the moon against his tables.

1. Clouds
2. Incapacitated
3. Equipment failure

Any of those could answer this, I think.  A fair question may be 'what is the most likely?'

Clouds were known in at least one region where the flight actually was, for at least some period of time - that was reported.  We don't have any specific report of Noonan not being able to 'shoot' due to clouds, so uncertain.

Incapacition due to... well, theories about 'personnel fitness' have abounded, and I won't be unfair to Fred who cannot defend himself; it is enough that it is a possibility for any number of reasons - including things like sudden, unexpected turbulance catching a tall fellow off-guard while out of his seat or something.

Equipment failure - always a possibility, although I believe someone posted recently how to check an octant if dropped, etc.  My guess is Fred knew the handling of such things so well and depended on them so much that it is not terribly likely - but back to the 'odds' - it can have happened, of course.

Regarding 3 didn't Fred have a second device or 'preventer'?

He may have - and if so, it tends to diminish the chances of celestial nav failure due to an instrument problem, of course.

I've always believed that is likely because of what he wrote about his pioneering experience with Pan Am.  Others disagree.  I don't know that we can ever prove that he did.  There is some history, by his own writing I believe, of having a marine sextant modified with a level for aerial application - and use as a 'preventer'.  I remain highly intrigued by the sextant box found with the skeletal remains at Gardner and the possibility that it relates.  Mr. Gaddy (I guess the famous navigator) examined it and proclaimed it to not be an aviation instrument box of that era, I believe - but perhaps he would have expected an octant and not considered that Noonan may have had a trusty marine device - perhaps old, but modified for aviation use. 

I do not know - nor may it even be knowable, unless we find a wreck full of junk one day...

Where would he sit to take the fix.  If it was up front AE could prod him awake.  If its behind the cockpit he could of been asleep nursing a bottle of Benedictine

He could have been up front or in the rear.  I've read somewhere that up front would be more likely for the sunrise position shot - and that clouds would have been a problem there none-the-less.  Also I believe it's been suggested that 'up front' would be desirable for spotting the island.

In the back, well, many things are possible... but I'd rather speculate about a banged head during turbulence than to throw even a possible aspersion on Fred about the alcohol.  We really don't know how real that was or what problems it really created in his life, I don't think. 

I don't recall any mention of sobriety being a problem during duty hours.  We have the cryptic Gore Vidal statement about Earhart's code for Noonan drunkeness being 'personnel problems' - but as I've been admonished before, that is extremely anecdotal - and she herself suffered stomach and sinus problems that may have been bad at times.  It could also relate to a dirth of local support on certain days or hours for all we know.

What's the most likely?  I'm cloud watching a bit on that... darn things can be everywhere when you don't need them to be, and Earhart obviously lacked the one thing she really needed the most for her best shot at an approach to Howland: competent RDF capability.

[JNev]

Even IF clouds are a problem couldn't they fly above the clouds?

 

At times, perhaps - but did they possibly commit to a lower altitude when they believed Howland was close, and perhaps lose the overhead sky?

Earhart reported flying around looking for the island at 1000 feet.  It would cost gas to climb again; she apparently believed she was 'on' Howland at that point.  I don't recall another altitude report with her subsequent transmissions, including the last one regarding 'on the line' - she may well have still been at 1000 feet trying to visually acquire while flying up, then down the line (I am biased toward believing that is the logical sequence 'on the line' - 'up' by 337 for a gas reserves-permitting time, thence 'down' by 157 until landfall).

For now, I keep roaming back to the 'clouds' - and where the clouds were.  It also suddenly comes to me that it is all the more tragic that Earhart could not get weather from the Itasca due to her radio ills - sky conditions in various sectors might well have become a vital clue in working with shipside observers to develop a better idea of where the flight was.  I guess that is something we can never really know, but it is sad to realize that there may be yet another tragedy within the radio debacle.

Don't get me wrong, either - I'm not saying Gardner could not have happened, but I think the discussion is worth having.  Nothing sharpens the chase like a bit of challenge on thought, IMHO, lest the mind gets too eager to see what it wishes to see.

[Scott C. Mitchell]

John Ousterhout’s correlation of AE’s radio broadcasts and likely navigation choices was fascinating.  The notion of cockpit conversations (maybe even shouting matches) gives this episode an even more human and poignant element.  I have been reading an excellent book called Deep Survival, by Laurence Gonzales.  In it, he offers the neuroscience studies of what happens when people find themselves in a state of being lost.  The hippocampus in the brain is apparently the epicenter of the effort to construct an analog world of a sense of place, a “spacial reference map” in the mind.  New data leads to an ongoing process of “remapping”.  However, stress interferes with this process, and what you end up with is a distorted sense of place and position.  It is this distortion that leads to faulty “seat of the pants” navigation decisions, where one’s hunches override the real data offered by instrumentation and calculation.  With an increasing sense of being lost, of not seeing any landmarks, the amygdala in the brain drives an unconscious urge to “get to a safe place fast”, and this sense also contributes to overriding mental calculation.  (Responses from the amygdala always override neocortex “logical thinking”.  When you jump at the snake in the path, you’re not “thinking” of the danger; instead, your amygdala has sent a direct overiding order to your body: “Jump!”  You do it without thinking.)  When lost person’s  mental map no longer offers a path to safety, in fact is shown to be completely unreliable, panic sets in, and this obliterates any remaining sense of orientation.  For people lost in the wilderness, there’s even a name for it: “woods shock”.   Doubt and second-guessing and almost random choices follow next.  In the case of AE and FN, add to that stress the exhaustion after almost 20 hours of flying.  And with two individuals in the aircraft, you could have two opposing disoriented “spacial reference maps” at odds with each other: AE, with her amygdala-driven seat-of-the-pants urge to head in a direction she “feels” is right, and FN, who can’t believe they are as off course as his spotty navigational opportunities suggest.  So at the point of “We must be on you but cannot see you” comment, what followed may not have been a cognitive flight plan to most efficiently search for Howland (box search, etc.), but fruitless low-altitude back-and-forth flying and cloud-shadow chasing.  After this, the panic-impulse from the amygdala would be to lunge forward, toward the next perceived “place of safety” – the Phoenix islands.  So there could be psychological as well as navigational impetus in that direction.

Scott Mitchell
Tighar #3292

[Ric Gillespie]

We don't need Gary LaPook.  We seem to be able generate pages upon pages of pointless postings without his help. 

- Dead reckoning all the way from Lae involves around 2556 statute miles +/-, so a 10% error would be in that case around 230 nautical miles or 260 statute miles.  That would be the expected outcome if the flight had had to slog all the way from Lae to Howland in clouds, without benefit of fixes along the way. 

DR accuracy depends almost entirely upon the pilot or navigator's knowledge of, and correction for, winds aloft.  We know that the forecast Earhart received was a wild-ass guess by a meteorologist in Hawaii who had no current data to work with.  The only report we have of winds aloft anywhere is at Howland at noon on July 2 and it only goes to 2,650 feet because they lost sight of the weather ballon in the clouds.  We have no information about winds aloft anywhere along the route.  Earhart's one mention of wind in a position report is ambiguous and useless.  Blithely postulating a 10% error for DR in the Earhart case is classic LaPook-style assignment of probability in the absence of facts.

- The flight had a fix at 0718 Z near Nikumanu Island - still 1700 SM from there to Howland; expected uncertainty from there to Howland would be about 170 SM;
- Then we believe they may have seen the Ontario at 1030 Z - about 1270 SM from Howland; uncertainty by the time of arrival at the 1912 Z call would then have been only about 127 SM;
- Then the flight passed Nauru at about 1130 Z - about 1143 SM remaining to Howland; uncertainty drops to 115 SM by the time they would reach the call point at 1912 Z;
- We believe they flew over Tabituea - only 613 SM remaining to Howland; that drops the uncertainty by dead reckoning to about 61 SM.

A house of cards.  "Expected uncertainties" based upon a totally bogus assumption and bad "facts."  Earhart probably saw SS Myrtlebank, not USS Ontario, but nobody knows for sure.  There is no evidence that Earhart knew either.  All she said was "Ship in sight ahead."
Nobody knows how far the flight may have been off course when it passed Nauru and the 1940 anecdotal story of an airplane being heard high over Tabituea can hardly be called a "fix."

All numbers, all probabilities, all based on what we know of Dead Reckoning.

And absolutely worthless.

  I believe the Nauru passage is reliable as a report from Earhart, but if I am wrong I will happily stand corrected.  There are other fixes I believe, so one need only do the basic math from those points and YMMV, of course - but odds is odds.

Earhart never made any reference to Nauru.  Earhart is known to have transmitted only two lat/long positions - one a little over 5 hours after departure which is clearly wrong and probably a transcription error; and the other two hours later which indicates a position 21 nautical miles NE of Nukumanu Island but she made no reference to seeing an island.

So which was missing, Fred?  Or the moon?

All the discussion of the moon ignores the fact that Earhart said they were "running on line north and south."  Why would they be doing that if Fred had shot the moon and had a fix that told him whether they were north or south of Howland?

Now I start thinking of Earhart's mention of clouds - and clouds to the north and west of Howland.

References to clouds to the north and west of Howland only turned up days later when Itasca was trying to explain why Earhart didn't see Howland.  There's no evidence that cloud conditions north and west of Howland on the morning of July 2 were any different than around Howland.

I won't continue.  My point is that facts count.

[Ted G Campbell]

All,

Has it occurred to anyone that once AE found she couldn’t DF into Howland, the fact they couldn’t see the island and the possibility that this was all due to FN not getting his sun, moon, etc. shots, FN would have instructed AE to gain/loose altitude so he could get a “shot” of something sometime before their arrival at Howland?

Seems to me that upon the approach to Howland and no one is answering your calls, you can’t see the island and the DF doesn’t seem to work, FN would have been busy trying everything under the sun (no pun intended) to determine his exact location.   The fact that he was convinced they were on the LOP tells me FN did obtain some type of “shot” just before entering an approach to Howland i.e. overcoming minimal observation impairments.

However, if there were a constant SE wind on their track you Nav. gurus should be able to determine about how far off FN would have been given his last minute sun shot – providing AE followed FN instructions to gain/loose altitude so he could perform his calculations.  The calculations I am looking for here is how far off track could they have been before the last minute shot would have indicated a significant error in a straight-in approach.  e.g. just my guess, 10 degrees off  on a sun shot, wouldn’t have been enough to cause concern with respect to the original flight plan but in reality would have put them x miles off course to the south.

Also, my guess is that FN told AE to first fly North on the LOP so he could get a sun shot out of the rear window at his nav. table – where all his charts and tables were – as opposed to being up front where he didn’t have the ready resources available to him to figure out where exactly they were. 

The other puzzle in these last minute approach snafus is was FN aware of the lack of communications between AE and Howland/Itasca?  Remember FN was passing notes via bamboo poles to AE, thus suggesting no intercom between the two, and I wonder if FN could even hear AE’s transmissions and/or land responses.  Was FN totally in the dark as to what was going on with regard to communications between the plane and ground support?  Did FN, once getting information from AE – not sure how this was accomplished, did she have a short bamboo pole? – decide he better get up front to look for an alterative landing spot?

Speculation, speculation I know.

Ted Campbell

[JNev]

We don't need Gary LaPook.  We seem to be able generate pages upon pages of pointless postings without his help. 

To split up my pointlessness, I’ll respond in two separate posts as I sense two areas of concern.

First -

There’s always the ‘red X’ of oblivion…

With all due respect, I think our friend Gary has suffered enough in absentia – we both realize by now he’s not going to appear to defend himself.  It really only makes TIGHAR look smaller than it should, don’t you think?  I plead for a kinder tone, even should a rebuke be believed in order. 

If you find my writing offensive, blame me personally.  It is mine, from my own understandings gained here and elsewhere.  I do not 'proxy' or 'ghost write' for others.  Whether I have drawn from Gary’s or other’s knowledge is my concern alone.  While I have pled for kindness in tone, it is for the benefit of the greater community and the sake of TIGHAR's image - personally I have fairly thick skin.

It was not my intent that any of this be taken on the chin or as ruinous to the Niku hypothesis, but to stir a bit of new thought and discussion.  IMHO as a member, this forum could use more vigorous exchange.  Sometimes it seems to be a shadow of what we had when I first came here.  I realize that promotional ideas are vital - but reason is also dependent on honest expression and questioning of substance.  I disagree that I have commented or questioned without reason.

I am not a ‘TIGHAR hater’ any more than a number of others who seem to share my view.  I do want to help ensure that the quality of exchange is fitting of what TIGHAR proclaims herself to be as a nonprofit educational institution.  My belief is TIGHAR would be endangered more by a slip in the quality of debate than lawsuits and spoiling trolls, and I for one refuse to fail you or TIGHAR by blowing smoke for comfort.  At times I may have done so unwittingly, but this isn't a good place for the unwitting; I've grown wiser and gained wits as a process of studying here and intend to be more scholarly and offer more of substance when I can. 

Please do consider that I was once accused of ‘being safe by hanging back’ when all I was really doing was challenging the implied near ‘certainty’ of certain things.  Nonsense – facts do count, and the fact is, we dwell for now in a highly ‘prospective’ environment where much remains unproven and highly speculative and circumstantial.  But that doesn't mean 'hopeless' or 'impossible' to me - and actually, I’ve spoken a number of times in this place of the need to accept risk if one would search.  The emerging Niku VIII plan is a grand example of that grit.  I believe if the Niku hypothesis is to be proven this kind of accepted risk and effort is what it will take, so I respectfully disagree with your analysis and judgment of my comments as ‘pointless’. 

That said, many discussions in a venue like this may well deserve a double-edged ‘hanging back’ question now and then.  But we do live with probabilities - and many of them remain subjective as to our personal judgment for now, so there will be differences - even when we agree to search in a certain place.  I’ll leave it at that. 

Thanks for this place and the opportunity to explore this mystery.

[JNev]

Second –

We don't need Gary LaPook.  We seem to be able generate pages upon pages of pointless postings without his help. 

Ric, you’ve made an assertion as to where the flight would have arrived on the LOP to fit a Gardner arrival scenario - some 230 NM south of Howland.  Your assertion of that particular was something I'd not realized before and it drove me into the navigation issues more deeply.  I learned a few things I’d not grasped before, that’s all.

For one, arrival at that point does not look like a given considering the things I have outlined – which I believe are well validated.  It could have happened, I don't doubt that.

As much as I admire and have enthusiasm for the Gardner arrival hypothesis and support exploring it, I still believe that other possibilities remain until irrefutable evidence is found.  That is of course just MHO, your's and other’s MMV from one end of the spectrum to the other, no foul from here.  That position is not meant as disloyal to anyone or any reasonable idea, merely that I see the mystery as remaining potentially broader than some others might.  As to the strength of the evidence found at Niku, IMHO the sheet metal article 2-2-V-1 and the plexiglass shard are among the most promising and interesting - that may well be because of my background, but if those could ever be tied to the Electra then we'd certainly have a concrete case IMO.

So -

- Dead reckoning all the way from Lae involves around 2556 statute miles +/-, so a 10% error would be in that case around 230 nautical miles or 260 statute miles.  That would be the expected outcome if the flight had had to slog all the way from Lae to Howland in clouds, without benefit of fixes along the way. 

Ric: DR accuracy depends almost entirely upon the pilot or navigator's knowledge of, and correction for, winds aloft.  We know that the forecast Earhart received was a wild-ass guess by a meteorologist in Hawaii who had no current data to work with.  The only report we have of winds aloft anywhere is at Howland at noon on July 2 and it only goes to 2,650 feet because they lost sight of the weather balloon in the clouds.

Nauru reported winds that were consistent with the forecast.  Itasca reported ENE at 30 on July 1st and at 22 the day earlier.  That does not tell all but it’s better than a WAG.  The weather was settled so far as we can tell, so it is reasonable that no big changes in the winds should have been the case.  Agreed – we cannot know for certain.  But, Noonan was equipped to determine drift for himself by computation between fixes with his drift meter, so an amendment of forecast information was likely possible aboard the plane. 

Ric: We have no information about winds aloft anywhere along the route.  Earhart's one mention of wind in a position report is ambiguous and useless.  Blithely postulating a 10% error for DR in the Earhart case is classic LaPook-style assignment of probability in the absence of facts.

I believe that Noonan did determine winds aloft at 23 knots of wind by himself at one reported point – if that relates to the ‘ambiguous and useless’ report, as you have declared it, then perhaps our MMV, but honestly and respectfully so.

Blithely?  LaPook style?  It’s your forum, present it as you will.

I recall the navigational assumptions from my own training, Ric.  I’m not dependent on Gary for advice on navigation.  I have a living uncle who is a Georgia Tech graduate (engineering) and who served in the navy as a navigator aboard the navy’s version of the B-17, flying out of Hawaii for long distances over open water after WWII.  Another uncle (now deceased) was a WWII carrier pilot and instructor who retired in 1968 and shared these things.  A brother was a career naval aviator who retired as a navy captain.  His aviation and sea experience were great.  I grew up with them sharing a great deal I continue to explore ideas with the surviving uncle and brother.  I also had a great flight instructor who was my first employer and in earlier years an early hurricane hunter pilot in the navy (P2V Neptunes, which he said ‘leaked a lot’ - they drove through the weather in those days, and later, Super-Connies - which could get above some of the weather).  He ‘shot the stars’ for fun quite well, and sailed extensively in the Pacific by DR and celestial means with another former instructor of mine - a fellow who left my hometown to go to work for a tiny outfit in it's early days as a Falcon captain: Federal Express.  Both of those gents knew a great deal of navigation and had lots of experience with open-water sailing and flying.  The first gent mentioned earned a bronze star flying a Catalina to pick up downed airmen in the open Pacific under Japanese fire and returning them to a remote island, safely. 

Great teachers, all.

Admittedly we cannot know all the facts – but the purpose of probabilities is to use them to consider and weigh the reasonable possibilities as best we might.  Benjamin Disraeli had a point – but as inconvenient as they can be, the numbers do speak.  YMMV, of course.

- The flight had a fix at 0718 Z near Nikumanu Island - still 1700 SM from there to Howland; expected uncertainty from there to Howland would be about 170 SM;
- Then we believe they may have seen the Ontario at 1030 Z - about 1270 SM from Howland; uncertainty by the time of arrival at the 1912 Z call would then have been only about 127 SM;
- Then the flight passed Nauru at about 1130 Z - about 1143 SM remaining to Howland; uncertainty drops to 115 SM by the time they would reach the call point at 1912 Z;
- We believe they flew over Tabituea - only 613 SM remaining to Howland; that drops the uncertainty by dead reckoning to about 61 SM.

Ric: A house of cards.  "Expected uncertainties" based upon a totally bogus assumption and bad "facts."  Earhart probably saw SS Myrtlebank, not USS Ontario, but nobody knows for sure.  There is no evidence that Earhart knew either.  All she said was "Ship in sight ahead."  Nobody knows how far the flight may have been off course when it passed Nauru and the 1940 anecdotal story of an airplane being heard high over Tabituea can hardly be called a "fix."

“House of cards” is a bit strong IMO.  I did not depend on all those points to make my point, and really only 'played one card' – I used the reasonably reliable Nikumanu fix.  BTW, that actually provides more margin of error to support your own supposition of a LOP-fall 230 NM south of Howland because of the fairly extreme distance involved – potential error is directly proportional.  Can you really validate the ‘assumption’ as ‘bogus’?

Consider - if Earhart saw SS Myrtlebank, and not USS Ontario, wouldn’t that put her around 25 NM further to the north of course than your extreme southerly arrival (230 NM south of Howland) would suggest?  Had the flight been 25 miles further north at that point then the likelihood of arriving 230 NM south is diminished accordingly.  That seems to inject a far more radical assumption of navigational ambiguity than I have used. 

-   It was around 990 NM from Myrtlebank I believe – which would make LOP-fall of 230NM south of Howland in the range of a 26% deviation error – more than 5 standard deviations (there go those pesky numbers again, I hear you Benjamin…). 
-   BTW, I goofed around with my old E-6 a bit – if the plane did overfly Myrtlebank and then wind up 230 NM south of Howland, then it must have hit sudden northerly winds (‘from the left’) of around 56 knots.  That event would not just rob the flight of a reasonable arrival point, but fuel as well – that’s a lot of crabbing (there goes some of my fuel for running the generator on the reef).  BTW, yes – I checked this with Gary, as a matter of fact, via pleasant email.  I am not a ‘TIGHAR hater’, nor is Gary, IMHO - and I am NOT a member of any other forum on Earhart, just to be clear.

I said “we believe” the flight went over Tabituea – clearly, we don’t “know” and I never “assumed” that.  We have a latter-day “anecdotal” report of a flyover heard, as you said. Perhaps I should have elaborated further in that regard, much as we have admittedly “anecdotal” evidence from Betty Klenck, Dana Randolph, Emily Sukuli, et al, etc.  Respected, but of course we can only take those things for what they are – and accept the ‘odds’, or not. 

Notice again that I did not use Tabituea or any other place in my actual navigational assumption other than Nikumanu.  That actually gave you the maximum benefit of DR error rooted in that fix, some 1700 miles before Howland.  For the sake of illustration I put the flight in the soup from Nikumanu only by using the values I gave.  I also believe Nikumanu is fair as a ‘fix’ because of the position report there – and it is conservatively favorable to your own presumption of a large deviation upon arrival at the LOP. 

All numbers, all probabilities, all based on what we know of Dead Reckoning.

Ric: And absolutely worthless.

Then without any idea as to probabilities, we should have no clue as to where the flight ended – including arriving on the LOP some 230 NM south of Howland or so.  Without probabilities, things like artifacts found in at least somewhat ambiguous settings have no value.  If nothing can be considered probable then we have a theory in a vacuum – one among a possibly infinite many, including that of a LOP flight to Gardner – and voila, lots of circumstantial if intriguing evidence.  Possible.  Perhaps ‘probable’ – within this universe of chaos.  Which I have never called ‘worthless’, BTW.  Probabilities exist whether convenient or not, that's all.

I believe the Nauru passage is reliable as a report from Earhart, but if I am wrong I will happily stand corrected.  There are other fixes I believe, so one need only do the basic math from those points and YMMV, of course - but odds is odds.

Ric: Earhart never made any reference to Nauru.

Thank you, I stand happily corrected, as I had offered.

Ric: Earhart is known to have transmitted only two lat/long positions - one a little over 5 hours after departure which is clearly wrong and probably a transcription error; and the other two hours later which indicates a position 21 nautical miles NE of Nukumanu Island but she made no reference to seeing an island.

True –

Earhart did not report seeing land at Nikumanu, but the position reported seems to have been approximately 12 NM west of the western end of Nukumanu, not NE (yes, I happen to buy Gary’s take on this – YMMV, of course).  Gary has made a convincing case to me that this is often gotten wrong variously because they use the published coordinates of the island, which are actually for the southeast point of the island – which is 12 NM long east to west.  If we differ, it is an honest difference of our interpretations of the data from that time. 

The root of this difference lies in a difference of interpretation - Gary has said he believes that Jacobson's interpretation of the minutes of the coordinates in decimal form is incorrect.  I had puzzled over that myself and having reviewed it with Gary, do favor his view of it – YMMV, of course.  That is no aspersion on Jacobson, a fine and smart man.

-   The 0718 report was “4.33 south, 159.7 east” – but it has been shown to me from other parts of the Chatter report that his usage of the decimal (".") is a separator between degrees and minutes – which I had come to believe myself, actually.
-   That, of course, is an opinion, but as such, my belief is that the correct interpretation would be 4° 33' South, 159° 07' East whereas Jacobson seems to have interpreted this as 4.33° South, 159.7° East. 
-   It has been pointed out to me that pilots and navigators of the day (and still, so far as I know, lest GPS gadgetry, etc.) used minutes and not decimal degrees for coordinates. 
-   That is not an attack on Jacobson or TIGHAR or any who agree with him, it is merely a differing opinion based on navigational experience from a few who know it better than I do, and I do know a bit of it too.

So which was missing, Fred?  Or the moon?

Ric: All the discussion of the moon ignores the fact that Earhart said they were "running on line north and south."  Why would they be doing that if Fred had shot the moon and had a fix that told him whether they were north or south of Howland?

That is kind of the point - if Fred could see the moon, why not shoot it and determine N-S position?  The moon had nothing to do with establishing a LOP and then running by DR ‘up and down’ the line, of course.  The point is they could have done so under cloud cover - and we really don't know how reliable even the LOP call was: it could have been tragically and inaccurately estimated by DR.  In fact, I happen to believe the LOP itself is  questionable – it is not clear that Fred could have gotten a clear sunrise shot with the clouds that may have been about.  The 'clouds' are the potential spoiler here, IMO.  Next is 'where would they have been to have been limited by cloud cover - did Thompson have a point?'

But, further to my point above – “where was Noonan? Where was the moon?”  Your own presumption of an arrival 230 NM south of Howland raised the point for me, which I had not considered well before.  Had they the moon, they well should have been able to determine how far south along the LOP they were; if they believed in the LOP, which apparently they did, then it could have become a simple choice as to 'which way Howland'.

So, ‘the big question’ remains, for me at least – “where was the moon”…

Now I start thinking of Earhart's mention of clouds - and clouds to the north and west of Howland.

Ric: References to clouds to the north and west of Howland only turned up days later when Itasca was trying to explain why Earhart didn't see Howland.

Was Commander Thompson lying?  See pages 5 and 6 therein for the 2 July, 1937 entry.  His early search actions were largely based on this observation.

Ric: There's no evidence that cloud conditions north and west of Howland on the morning of July 2 were any different than around Howland.

To the contrary -

From Thompson’s ‘Cruise Report’, for 2 July, 1937 –

“During the last half hour prior to getting underway an estimate of the situation was made based upon the following facts and assumptions:

“FACTS”
(J. Neville comment – this can be read on page 6 of 12 via the above link, so I’ll only quote the cloud stuff; the visibility comments are, however, also interesting – and suggest the flight might not have gotten within 30 or 40 miles, I believe).

(c) Visibility north and west of Howland excellent to horizon but beyond that continuous banks of heavy cumulus clouds.

(d) Plane transmissions had indicated that dead reckoning distance had been accomplished.
…
(h) Stellar navigating possibilities, south and east of Howland and close to Howland, were excellent throughout the night.

Now, where was the moon?

Ric: I won't continue.  My point is that facts count.

My point exactly. 

It is not my intent to 'tear down' anything here.  My intent is merely to sharpen the sense of the quest for those of us who seek to learn and to work all the problems inherent in testing the Niku hypothesis, that's all.  Others MMV, but in my view we are stronger when we openly and thoughtfully challenge points of dependency - if that is what a 230 NM arrival south of Howland on the LOP is (I took it to be, but perhaps over read the matter?). 

As an example, recall that a found 'navigator's bookcase' once generated a great deal of interest and speculation here about what Fred might have had installed in the Electra for his stuff.  Finding it to be likely not related did not kill the search - it merely meant sharpening focus elsewhere and continuing to work with the other possibilities.

Thanks for this place and the opportunity to explore among all that TIGHAR has provided.

[richie conroy]

Whats your point Jeff or should i say Gary / Malcolm /  Norbert Bert Ernie ?

Anyone else's name is not worth a mention,

Can you provide a link to your websites to assess my self if there is anything to it ? By the way i have been through them all an it is only yours Truly that has provided most adequate info  that is upto you to decide

xx

[richie conroy]

No Chris

The difference was, We are able to decipher truth from what ever goes, while to most i just go along with what goes,  I beg to differ i have numerous UN answered questions i.e why no other voice messages straight after we are on the line ? we only seem to hear what is compatible with Tighar hypothesis But YOU like me don't need that volume of critical analyze to be convinced by Tighar's work if you have had a change of heart so be it, am not here to convince you only to remind you every new scenario you can create Tighar have fended off equal claims   of that hypothesis

 Richie