Forum artHighlights From the Forum

September 24 through 30, 2000
(page 2 of 2)

(click on the number to go directly to that message)
18 Tides and Storms John Pratt
19 Takeoff Distance Tom Roberts
20 Something New Michael Holt
21 The Notebook Mike Everette
22 "Betty" in St. Pete Don Neumann
23 Post-Loss Receptions Hue Miller
24 Post-Loss Receptions Hue Miller
25 Films & Songs in the Notebook Ric Gillespie

Message: 18

Tides and Storms

Date: 9/30/00
From: John Pratt


I note the interest in tides, especially related to the hypothesis that AE and FN landed on the coral flats on the west side of the island. I can see three, maybe four questions to be answered with tide information.

  1. The water depth on the flats at the expected arrival time.
  2. Correlation between credible transmissions and low tide.
  3. Correlation of the Navy visit and high tide.
The first three can be addressed with a simple tide table. The fourth requires a different type of tide information, a "Tide Chart". Much rarer, this would be a flow map for the limited area. Usually these are available only for areas of considerable economic importance (such as New York Harbor).
4. Perhaps even breakup rate and dispersal direction for the wreckage.
This is recognized in the current FAQ entry and the implication that enough people ask to justify FAQ production isn’t lost on me:

Is it possible to know what the tide was like at Nikumaroro when Earhart may have arrived there?

No, it isn’t. There are no tide tables available for Nikumaroro now, let alone 1937, so the information would have to be extrapolated from Samoa (700 NM away) or Kanton (200 NM away). Unfortunately, such extrapolations are not reliable. Add to that the distance in time, and the fact that it only takes a small error to add up to six hours over 62 years, and it would be quite possible to be exactly opposite the reality. Even on the 1999 expedition, extrapolation from current tide tables available for Samoa turned out to be off by a factor of approximately 50%--a three hour error.

So I tried the people who do tide predictions: NOAA.

They have an excellent FAQ, providing enough background for even me.

Basically, they offer tide tables for any date, even 1937, for anyplace they have the "Harmonic Constants" used to calculate the Tide Tables.

So it doesn’t hurt to ask:

> I have an active fantasy life, so let me share some ideas:
> I. Existing NOS data allows calculation of a tide table for that place and
> time.
> II. A future expedition could make tide observations which establish the
> relationship to well-known locations in the region. From this a tide table
> for the earlier date could be calculated.
> III. A future expedition installs a solar-powered tide monitor, leading to
> establishment of a sufficient set of harmonic constants for historical
> predictions of sixty years ago.
> Is there any reality in any of these fantasies?


I. NOS does not have tidal data for Gardner Island. Therefore calculations cannot be generated for the specific location.

For the immediate:

We cannot provide information specific to "Gardner Island", but we do have information for "Canton Island" in the Phoenix Island chain. Attached you will find tide predictions for Canton Island during July 1937.

>> II - III. Assuming observational data does become available at some point
>> in the future; this could be relatively short term data to allow the
>> establishment of a secondary station, or harmonic constituents.
>> This would allow predictions
>> to be generated for the location. We can generate predictions for past dates.
> What would this observational data consist of?
> What sample size is necessary?
> Is there a sampling frequency or methodology requirement?

The water level gauges we use record observations on a 6-minute interval. There would have to be continuous data taken for at least 1 month. With a 29 day data set, we can generate a basic set of harmonic constituents. However, with that short a data set we cannot directly calculate some of the constituents, they have to be calculated from the few which are observed. Data sets of at least 6 months allow more constituents to be directly derived, and a year provides all but the very long term constituents. Generally, the longer the data set, the more accurate the constituents.

For more information on the exact requirements for data, please contact Steve Lyles (

> 1. Your web site gave a very clear explanation of the effects of possible
> local variations. Many of those sources (such as bay-locations and runoff)
> do not seem to apply here. Are there any outstanding reliability issues
> that apply to the edge of an isolated desert island?

There are other effects on tides which may affect islands, most of these are concerned with the effects of continental shelves, restricted passages, etc. In general, these have a significantly smaller effect than the situations of stations located inside bays or up rivers.

> 2. IF data were obtained to allow calculations
> AND subsequent (modern) observations were obtained and compared with
> those calculations (sample size for valid test unspecified)
> AND the the calculations and observations were considered to agree (test
> unspecified)
> THEN Could the calculations be considered validated and applied to the
> historical situation? (in other words is a solution expected to be stable
> over sixty years given minimal environmental change?)

With a set of constituents, reguardless of when the data is obtained, we can generate predictions for virtually any date. Tidal constituents don’t change all that much over time. Most of the changes would be due to evnironmental changes such as shoaling or errosion, and artificial changes such as jetties, dredging, etc.

> 3. IF data were obtained to allow calculations
> AND subsequent (modern) observations were obtained and compared with
> those calculations
> AND the the calculations and observations were considered to NOT agree
> THEN Could the calculations be adjusted to match the observations and
> tested for validation as above? (in other words, would deviations be of a
> "systematic nature" or is the variation a "statistical" process?)

Part of our process to generate constituents is to use the derived constituents to generate predictions for the same time period as the data. The two are then compared to verify the accuracy of the predictions and any adjustments are made at that time.

> Canton Island is roughly 200 NM away.
> Do you have a correction formula to convert from Canton to Gardner?
> Do you have any way to estimate the accuracy of such a conversion?
> Can 3. above be applied to increase accuracy?

No, we cannot convert from Canton to Gardner. That would require data at Gardner. If we had that information we could calculate it directly.

Much of this was confirmatory, but I see a possible way forward. It’s an instrumentation problem, a "Long Term Monitor (LTM?) project (unless NOS would loan a portable data acquisition station). A month of data might be possible.


One of the Tide Charts mentioned above is almost certainly out of the question, but I see two cases in the record where there are storms in early July, 1937.

Could the storm surge waves be more important than ordinary tides for aircraft breakup and dispersion? It seems likely.

Could the wind direction (and implied flow of water) provide a direction for that dispersion? It would be nice to investigate.

So I see two mention of storms: As early as 1138 on July 2 Itasca had suggested that the Navy send a patrol plane to assist in the search, but it was 1923 before a plane headed south under the command of Lt. W. W. "Sid" Harvey. He and his seven man crew would spend the next twenty-four hours and three minutes aloft only to land where they had started D forced to turn back barely three hundred miles from Howland by "extremely bad weather." (An Answering Wave.)

Another report of the same incident:

At 0700, the Patrol Plane reported her position at Latitude 6°35′ North, Longitude 172°00′ West, that the weather was extremely bad and that it was necessary for her to return to Pearl Harbor. (Friedell’s Report.)

And one I can’t find my way back to, a mention of Colorado experiencing heavy weather and seasick NROTC cadets on approach to the general search area.

Assuming that the PBY and Colorado experienceed the same storm, it must have been relatively large, a significant regional weather feature. Its intervention before the Colorado aerial search suggests that it could have inflicted enough damage on the Electra to prevent observation in the search on July 9.

Is there a record from the Itasca or the Swan logs? One would expect that those logs could provid wind speed and direction and sea state. Would they also mention the direction of local wave action?

John Pratt (Number not yet received)

From Ric

Thanks John. Good observations and good follow up. We’ve long wondered about the swells generated by that disturbance. Bob Brandenburg recently made the following observation:

The swell factor is interesting because even without wind, swells from great distances can impact the reef. The ships’ weather logs show that Colorado and Swan were both in the general neighborhood of Gardner for several days before the 9th of July (Colorado getting within about 60 miles, and Swan about 250 miles northeast). Both ships reported "moderate" and "heavy" swells with no sea (wind waves) from the east and northeast on the several days preceding the 9th.

Itasca,operating within roughly a 250-mile radius of Howland from 2 through 8 July also reported "moderate" and "heavy" swells from the east and northeast. So we have evidence of a generalized easterly swell system extending at least from Howland to Gardner. Swells rolling up on the reef during high tide would be a major threat to the aircraft. Waves generated by local winds added to the mix of forces. I’ve worked out the statistical distribution of wind speeds and the resultant generated wave heights, and there were periods of impressive waves (higher then 5 to 6 feet) coming ashore. Combined with swells and high tide, they created a lethal mix of forces pounding on the aircraft.

Message: 19

Re: Takeoff Distance

Date: 9/30/00
From: Tom Roberts

I’m a little late with this, but it still may be of interest.

A few months ago, Oscar Boswell posted a very impressive (at least to me) assessment of the Lockheed’s performance based on similarity with other comparable aircraft. However his "back of the envelope" calculation of the effect of a five-knot headwind on takeoff distance (~300 foot reduction) left me doubting, as usual.

So I tried something a little more rigorous, using Excel and Newton (F = m x a). First, I iterated on net thrust (constant engine thrust minus a constant friction force) and drag coefficient (secondary effect) until the 15,000 pound aircraft reached 100 mph over a distance of 2900 feet in still air. This takeoff run took approximately 38.8 seconds. Then I included the effect of a 5 knot (5.75 mph, 8.43 ft/sec) headwind. The aircraft reached 100 mph (airspeed) in 36.6 seconds over a distance of 2582 feet. The difference of 318 feet is in very close agreement with Oscar’s numbers.

LTM (Who never doubted Oscar was right)
Tom Roberts, #1956CE

From Ric

It’s never too late for good numbers.

Message: 20

Re: Something New

Date: 9/30/00
From: Michael Holt

Betty’s notebook tale is great, but -- as has been suggested -- why could it not be a local production? Was that not the era of live and local dramatic productions? Is there any reason to feel that station(s) did NOT create such programming?

If it’s possible to get some idea of the frequency Betty was monitoring, would it not be worthwhile to figure out what station(s) might have created a dramatization of the crash?

This kind of stuff makes me suspicious. If Betty’s notebook can survive a TIGHAR attack, it’s gotta be real.

LTM (who knows that Kato wasn’t Filipino until December 8, 1941)
Mike Holt

From Ric

The best argument against what Betty heard being a local production is that she heard it very sporadically on short wave over a period of more than three hours. Betty’s neighbor could not pick up the signal but did not have the large antenna array that Betty’s father had erected. If it was not the real thing it seems more likley that it was a fairly elaborate hoax perpetrated by someone at a considerable distance.

Message: 21

The Notebook

Date: 9/30/00
From: Mike Everette

This brings to mind the Orson Welles "War of the Worlds" broadcast on Halloween in 1938. Mass panic... it sounded "real" to people then; but to us nowdays who are more sophisticated, used to massive media coverage, when we listen to recordings of the broadcast it sounds hokey or quaint.

Before someone suggests that somewhere in the annals of network history there may be a record of any such dramatization of AE’s predicament, please recall that there were different networks in those days... records may be hard to come by.

Also, be assured that many local stations produced their own dramatic shows in the 30s... not all programming came from a network.

Whether this is real (we hope... dream??) or a hoax or a misinterpretation may be most difficult to prove.

Something that lends a modicum of credence: if "Betty" was into radio, airplanes et al, remember that a lot of a/c communications could be heard in those days on the average home "all wave" receiver, and many home sets had SW since FM was still in the future. People were crazy about radio then, like they are over computers today. Short wave radio was hi-tech in the 30s.

6210 KHz is very close to the 49-meter SW broadcast band (actually at that time it was in the midst of it). It would be logical for someone to be tuning the 6-MHz band and quite possibly stumble onto something.

This would be a longshot... a real "long haul" halfway around the world on 49 meters... but by some fluke it could be remotely possible. I have myself heard some strange things I’d have considered impossible. Never say NEVER regarding HF propagation... but I sincerely doubt it could ever happen on 3105.

LTM (who hears EVERYTHING) and
Mike E.

Message: 22

"Betty" in St. Pete

Date: 9/30/00
From: Don Neumann

Think you’ve got a phony here, in all of her recorded radio transmissions AE _never_ used her name, only her call sign.

How was ’Betty’ able to ...’get the impression’ ... that the ’man’ had a head injury or ...’gathered’... that they had crashed on land & were threatened by water, when, having trouble keeping up with the speaking speed of the person broadcasting (she was only jotting down a word or two, here & there) on an ordinary shortwave set, from a distant, unclear radio signal that was ...’fading in & out"..& was ’distorted’?

Perhaps you should question her first about just how many AE books she’s read over the past 63 years (some of her revelations sound very much like stuff I’ve read in some of those books), as I suspect you have here some very old recollections of a woman, who once was a 15 year old with a very active imagination, who may have simply appropriated for her own journal some of the subsequent details of the search for AE, including any newspaper accounts of the many radio messages allegedly heard by numerous operators during the time frame after the plane was presumed down & missing.

Even if you can’t track down any corroberating witnesses or family members, at this late date, maybe newspaper articles appearing in the local St. Pete press of that era, might provide some clue as to just how much information was reported in the ’local’ newspapers about the radio messages allegedly received after the plane was presumed down.

Should be interesting what the other Forum members have to say about your latest ’find’ & especially those from the ...’other net’..., many of whom insist that not only were the post-landing messages valid, but that they somehow pinpointed the location of the landing in the Marshalls or Saipan!

Don Neumann

From Ric

This kind of rush to judgement is exactly why we’re not releasing the full transcript until some real work has been done.

Message: 23

Post-Loss Receptions

Date: 9/30/00
From: Hue Miller

One complication to the post-loss receptions is the low tuning dial accuracy of radio receivers in use by hobbyists – or actually just about anyone – in 1937. IF you had once heard a station for sure on 3105 or 6210, and you marked your dial or noted the position, you could maybe reset the dial to within 5 kHz or a couple kHz at best, if you decided you needed to monitor on those frequencies again. If you hadn’t noted the calibration ahead of time, you might be within ± 10 kHz or even 20 kHz. With a lot of background noise, weak voice coming in & out of the noise, guesswork or imagination enters the picture.

Hue Miller

From Ric

So a one-time event makes more sense than if somebody claimed that they heard signals on different occasions?

Message: 24

Re: Post-Loss Receptions

Date: 9/30/00
From: Hue Miller

Ric asks:

> So a one-time event makes more sense than if somebody claimed that they heard
> signals on different occasions?

No, all I am suggesting, and I really don’t know the content of the various post-loss receptions, that people would have been easily tuned to some other frequency than 3105 or 6210, they could have been hearing some other kind of weak, fadey signals going on, and then their own imagination kicked in.

If they heard a signal and it was in fact nearby but not exactly 3105, for example, 3106, you would know for sure it was a hoax, as AE’s transmitter was crystal controlled and would not vary more than some few tens of Hz. (I see that her transmitter had a heating element next to each crystal ( "crystal oven" ) to regulate the crystal temperature and keep the frequency solid. Actually, unnecessary overkill in the world of AM equipment....) However.....most listeners then didn’t have any way to check the tuned station’s frequency that we cannot prove or disprove anything by this....

Hue Miller

From Ric

That’s right. Betty has no idea what frequency she was listening on. She was just tuning along the dial and stopped when she heard something interesting.

Message: 25

Films and Songs in Betty’s notebook

Date: 9/30/00
From: Ric Gillespie

Below are listed the films and songs referenced in Betty’s notebook in the order in which they appear. Determining the release date for each could help establish when the notebook was in use.

Page 23

Black Legion - Humphrey Bogart
God’s Country and the Woman - Beverly Roberts
Elephant Boy - Sabu
Swing High, Swing Low - Carol Lombard & Fred MacMurray
The Great O’Malley - Pat O’Brian
Maytime - Jeannette McDonald
A Day At The Races - Maureen O’Sullivan
A Woman of Glamour - Kent Taylor

Page 29

It Looks Like Rain (on Cherry Blossom Lane)
(lyrics written down)

Page 31 (numbers may refer to Hit Parade ranking?)

9. Love Bug Will Bite You
2. Carelessly
7. Where are you
6. Sweet Laylone
5. There’s a lull in my life
3. Never in a million years

(Then written below)

Merry-go-round broke down
Never in a Million Years
A sail boat in the Moonlight

Page 33

The Love Bug’ll Bite You
(lyrics written down)

Page 34

Where Are You
(lyrics written down)

Page 35

(lyrics written down)
Just a Quiet Eve

Page 36

There’s a Lull In My Life
(lyrics written down)

Page 37

A Sail Boat In The Moonlight
(lyrics written down)

Page 38

They Can’t take That Away From Me
(lyrics written down)

Page 41

It Looks Like Rain In Cherry
(lyrics written down)

Page 43

Where Are You?
(lyrics written down)

Page 44

Johnny One Note

Page 46

All God’s Children’s Got Rythum [sic]
(lyrics written down)

Page 47

Merry-go-round Broke Down
(lyrics written down)

Page 48

You Can’t Run Away From Love Tonite
(lyrics written down)


Return to Page One, Messages 1 -- 17.
Back to Highlights Archive list.

About TIGHAR Join TIGHAR TIGHAR Projects TIGHAR Publications Contract Services
The TIGHAR Store Blog TIGHAR Forum Contact TIGHAR TIGHAR Home

Copyright 2018 by TIGHAR, a non-profit foundation. No portion of the TIGHAR Website may be reproduced by xerographic, photographic, digital or any other means for any purpose. No portion of the TIGHAR Website may be stored in a retrieval system, copied, transmitted or transferred in any form or by any means, whether electronic, mechanical, digital, photographic, magnetic or otherwise, for any purpose without the express, written permission of TIGHAR. All rights reserved.

Contact us at:  •   Phone: 610.467.1937   •   JOIN NOW