I met Frank Stewart eight years ago after his family contacted TIGHAR. At the time he was in his 90s and living in the San Diego area. I was in LA doing research for the NBC News documentary on Earhart, so I was doubly pleased for the opportunity to go and talk with him.

"Pops" Stewart began his Coast Guard career as a raw recruit chasing prohibition-era rum runners out of Woods Hole, Mass and finished as the navigating officer of an assault transport in WWII. In the mid-30s he served as the Quartermaster of the cutter Itasca and sailed with her from Alaska to South America and all points in between -- including, of course, Hawaii and the Line Islands.

It turns out that he was also a very gifted photographer and took scores of pictures during his travels. He supplemented his pay by selling copies to his shipmates and the originals he kept in his scrapbook. I was shown the photos while a guest in Lt. Stewart's home and they kindly let me borrow the albums and use whichever pictures I wanted for the show.

The much debated photo of Itasca "making smoke" is from a series of photos taken during Itasca's visit to Howland in 1936. Lt. Stewart looked, but could not find any pictures that were taken during the Earhart search. I felt justified in suggesting the photos as illustration for the NBC show as they were as close as anyone could come to the real event (same ship, same island, different year).

However, the major source of confusion probably stems from a special file at the US National Archives. The folder contains pages from the original Itasca radio log that were saved by Leo Bellarts and later donated to the Archives. It also includes a couple of Frank Stewart's pictures -- one of which is the "smoking" photo, only labeled July 2, 1937.

The composition of the two prints are identical. There is no doubt in my mind that Bellarts got his copy of the picture from Frank Stewart. As for the date discrepancy, I chose to believe the one written down at the time by the photographer himself (i.e. 1936). IMHO, the photo of Itasca "making smoke" may be evocative, but it's not evidence.

LTM,
Russ

From Ric

Thanks Russ. Sounds like Bellarts may have bought a copy of the smoke photo from Stewart and later included it with his original radio log when he dontated his stuff to the National Archives. It also sounds like, if the National Archives photo has "July 2, 1937" written on the back, the date is incorrect and was added much later.

A War Department technical manual from 1941 contains this interesting caution about the A-5:

The main weakness of this instrument is that it employs the make-and-break type of bubble. This kind of bubble works well except that it requires extra time for adjusting before each series of observations and unless carelfully handled often gives trouble. The size of the bubble varies with temperature, and the vibration within the plane frequently causes the bubble to leave the level chamber. Until an observer acquires the knack of forming and adjusting the bubble, there exists an ever present chance of damaging the bubble assembly.

As to the precision of the A-5, the manual says a degree on its angle scale was "subdivided into twelve parts each reading 5' of arc."

The manual describes the A-5, the modified A-5, and the A-5A. The first two have no averager. But "Type A-5A has been further modified to incorporate a device designed to average automatically any eight altitude settings."

>I don't
>know what was in use at the time, but I would not have been surprised to
>find FJN lugging a nautical almanac and reduction tables.

Two Nautical Almanacs were listed in the Luke Field inventory. The 1937 American Ephemeris and Nautical Almanac tabulated the Moon's semi-diameter and horizontal parallax for 12-hour intervals, and measured the angles to the tenth of a second. The Moon's Right Ascension was given at one-hour intervals. Sidereal Time--the time equivalent of the Greenwich Hour Angle of Aries--was given daily.

The American Air Almanac was first put out in 1931, but was discontinued for lack of interest. The British came out with one in 1937, and the American almanac resumed shortly.

Two copies of Dreisonstok's "Navigation Tables for Mariners and Aviators" were also in the inventory. With them the navigator can compute height and azimuth. More convenient tables tabulating height and azimuth had been available for some time--"Altitudes and Azimuths" came out in 1919--and Noonan might well have had them along, too.

>.....the moon has always been a major celestial
>body. Not too awfully long before FJN, marine navigators were still
>checking their chronometers by the lunar distance method, and the Nautical
>Alamanc had tables of pre-computed lunar distances.

The "American Ephemeris and nautical Almanac" last tabulated Lunar Distances in 1911. The technique was still possible though without the tabulation--just more computation was necessary.

> I'd bet that moon data
>and sextant corrections were always included in the almanacs.

The moon data were; tables of sextant corrections for dip and refraction were in the Dreisonstok book.

>.....problems with.......some other aspect of the air nav process? Any good
>navigator would have used every means possible to fix or confirm his
>position
>
>.....the flight did not end as intended. If we attribute that to problems
>with navigation, then something seems not quite right with the picture.

There was a technique of pre-computing the expected heights of the body as seen along the LOP running through the destination. The heights would be graphed; a gradual curve would result. A rule was developed which would tell which side--in his case, the 157 course or the 337 course--the plane was on. Here is how it is described in, "Celestal Air Navigation," put out in 1941. The rule was similarly stated in Weems, Air Navigation.

Celestial landfalls need not necessarily be made by directing airplane defnitely to one side or the other of the objective. If the navigator is in doubt he should take up a course of azimuth +90 degrees or azimuth -90 degrees, whichever he believes to be correct, as soon as his observations intersect the curve. Then at intervals of about 5 minutes he takes at least three sets of observatons which will indicate whether objective lies ahead or behind by the rules that--

When course is azimuth +90 degrees, if altitudes plotted are--
Less than curve, destination is ahead.
Greater than curve, destination is behind.

When course is azimuth -90 degrees, if altitudes plotted are--
Less than curve, destination is behind.
Greater than curve, destination is ahead.

But this published rule is in error. It is not generally true; it is true only when the body's azimuth value is increasing (as when a body is passing south of the objective). Both the Moon and Sun were to pass north of Howland on July 2, and their azimuth values were decreasing. Application of the rule would give a wrong idea of direction from Howland along the LOP. But I don't know if Noonan was aware of the rule; if he was I suspect that he would have seen the danger of its being applied erroneously.

Mark Prange

From Ric

Holy Moley, the deeper we dig into this the more layers we find.

> Is there a way that Randy's solution (and others for that matter)
> could be reproduced visually with step-by-step explanations, and then
> distributed in the next TIGHAR Tracks? This whole issue is VERY confusing
> to most of us and I think a "show-and-tell" approach would better educate
> all of us.

Dennis, you have hit on the very problem that caused me to post my first few notes.

It quickly became obvious from the postings that only a very few people on the forum understood, even vaguely, what an LOP was. Postings showed they didn't understand how Noonan could get on the LOP or know he was on it. Consequently I tried to craft an explanation of that one issue without confusing everyone more with a lengthy course on celestial navigation. An attendant problem was how to do that knowing there were a couple of people who DID know something about celestial procedures. I decided to totally ignore the dead reckoning (DR) portion (which I announced) and to do that I had to sort of fudge. I stated in one of my first notes that I was only dealing with the LOP issue and that what I was writing was technically NOT accurate and that this was not the way it was really done. I did that solely to eliminate all the work going into shooting and plotting the fix to show ONLY what an LOP was and why Noonan could know when he was on it whether he was ON course, or north or south of it.

While working on this it dawned on me that the 67 degree azimuth might be significant information so I contacted a number of professionals in this regard. After giving the background I posited that the 67 degree sun azimuth might tell us something as to where in general Noonan might have been during a very small time and position envelope. I contacted the US Naval observatory, the British outfit that publishes their air almanac, the aerospace mechanics and engineering department at the University of Texas, a Canadian government observatory and several other professional astronomers. All believed it was a reasonable and possible experiment and mathematically workable -- my detractors to the contrary.

One forum member who is quite knowledgeable of celestial procedures instead of seeing the possibility got side tracked by his accurate knowledge that a sun azimuth is not actually shot but rather only the altitude. This is true to a certain extent. The azimuth is instead plotted using the sun shot and the almanac tables. It also is shot using an azimuth circle which was a standard technique long before this particular flight. It is also true that when shooting the particular body the azimuth will tell the navigator valuable information.

The azimuth of the sun, contrary to what has been suggested, does NOT remain the same all day. It changes with the observer's time and position. If you will look at the sun data for Howland's position for the morning of July 2, 1937 you will see that the azimuth varies throughout the day. If you pick a position dead on the island and look in the table for a particular time, say 07:00a you will find a certain azimuth. If instead you pick a place 60nm dead north and check the table you will NOT find the same azimuth. It varies from north to south.

If you will picture yourself that morning 180nm west of Howland on a true course of 77 degrees and you look at the sun and note its direction, then picture yourself three or four hundred miles either north or south of that course and it should be obvious that the sun MUST be at a slightly different angle.

The bottom line is that there may be enough information to GENERALLY place Noonan between sunrise and 8:45 local.

While going through this exercise it also dawned on me that everyone has pretty much made the loss a navigation issue. Noonan has been restricted to only a sun line. He has not been allowed to shoot the moon or planets all of which was available to him that morning. I have also checked and found the data WAS available in the charts of the day. There were even SOME air almanacs but what they contained for the relevant period I haven't determined. I have found that the British air almanac was available but only for the last quarter.

We have not allowed Noonan to use an azimuth circle that was standard for the day nor have we even given him the reasonable possibility of checking his drift and ground speed. ALL normal navigation techniques that ANY navigator would have used let alone a MASTER navigator with years of nautical and Pan Am experience. In our rush to explain why they missed Howland we have stripped Noonan of all his great experience and expertise and the tools of his trade. We have created an impossible task for Noonan in addition by exploiting all the known celestial errors and variables and somehow implying Noonan would either not have known of them or could not take them into consideration. Some have even made him incapacitated from drink.

And yet the possibility that our duo simply could not visually spot a tiny island among all the cloud shadows seems too far out to accept. Interesting.

Alan
#2329

Mark Prange wrote:

>Two copies of Dreisonstok's "Navigation Tables for Mariners and Aviators"
>were also in the inventory. With them the navigator can compute height and
>azimuth. More convenient tables tabulating height and azimuth had been
>available for some time--"Altitudes and Azimuths" came out in 1919--and
>Noonan might well have had them along, too.

Most of this navigation stuff is over my head, but for what it's worth, Noonan said, "I suppose you wonder which method I use for computation of observations. I use Dreisonstok exclusively. Probably another prejudice, but I have used it since it first became available in 1927 or 1928, and still prefer it." This was in his letter to Weems (May 11, 1935) written shortly after the first Clipper survey flight to Hawaii to test the first leg of the eventual route to Manila.

blue skies, Dreisonstok or otherwise, -jerry

Doug B asked about flight path, and I started looking at the 0718 GMT position report (I have never made sense of the 0519 report). I have always thought that this resulted from a full fix, since it includes the statement about windspeed (23 KT) which would seem to have been computed by comparing a DR position with a fix.

However, it turns out that at 0718 (if the time was correctly reported) at the position reported (4-20 S 159-42 E) the sun had not quite yet set (sunset at 0721). The moon had long since set at 0132 GMT. Bright stars or planets would not normally be visible at this time, and probably not for up to 30 more minutes. This leads me to believe that whatever this was, it was not a full (2 or more LOP) fix.

It could have been a running fix by crossing an earlier sun line with one a little before 0718, or it could have been a crossing of the DR track with a sun line before 0718, or just a plain DR position. But I don't see that it could have been a full fix. It is hard to say how accurate a position report it was.

The above data is for sea level elevation. Sunset at altitude will be later. USNO website: http://aa.usno.navy.mil/AA/data/

Regards,
Tom MM

From Ric

The position reported at 0718 GMT is roughly equivalent to the position of the Nukumanu Islands and it has been so frequently assumed that a sighting of those islands was the source of the fix that it is now taken as a "given" in Earhart lore.

Whether or not that assumption is correct, it is also true that Randy Jacobson has been able to determine from the records of the Oakland/Hawaii flight in March that Earhart's position reports never reflect her present position but are merely the most recent position information she has at the time of her scheduled transmission.

Well, this is great information thanks to Doug B and Mark P. I'll try to summarize/paraphrase what has surfaced so far.

1. Noonan probably used a Pioneer A5 bubble octant. Only later versions of the A5's had an averager --- these were not chronometric (for say, 2 minutes), but instead averaged any eight discrete sights automatically. If it was an early model, Noonan would have had to sight, read, record, and sight for a series of measurements and then manually average them or select the "best" from the series.

2. The A5's angular range was about -5 to +95 degrees, and read to 5 minutes of arc.

3. The main weakness of the instrument was its bubble type which was temperature and vibration sensitive. It required adjustment before a series of sights, and if misshandled, could give trouble. Until the observer acquired skill with the bubble adjustment, there was a risk of damaging the bubble assembly.

4. Two versions of the Nautical Almanac were on board on the first attempt. Moon data were tabulated. Note: An Air Almanac was published in Britian in 1937, and shortly afterward, in the US (again).

5. Applicable sight reduction tables were on board on the first attempt.

6. The AE/FN Electra did not have easy all around sight capability for the navigator. If the aircraft heading could not be briefly changed (for several minutes) to take sights, FN might have had to move from back to front or visa versa to get good cuts for a full fix. Sights might have had to be taken with the instrument pressed close to the glass and held at an angle.

Corrections-- additions?

Regards,
Tom MM

We may have been too hard on Capt Thompson afterall. Maybe various authors took excess liberty with his smoke plume and created a plume so that AE could have or should have seen it.

Capt Thompson sent an official report to the Commandant 14th Naval District that on 2 July he began laying "heavy smoke screen at daylight." Period. In this report he makes no further comments about the smoke's height, density, direction, or potential visibility, particularily later on that morning around 0843. Odd he uses "smoke screen" not "smoke plume".

And in all fairness, he added that (under assumptions) that the plane..."may have missed the smoke screen, ship or island visually due to their lying in the glare of the rising sun."

I can't figure out who wrote the report refered to in Lovell's book that the Itasca's smoke "plume" could have been seen 40 miles or more. Thus I think your're right that little smoke was visible around the critical time AE is circling and that the "visible smoke plume" should not be associated with Capt Thompson.

Is the ship's log (not the radio log) forthcoming in your book or has it been posted?

Radio Communications stuff. Did Goerner correctly quote Paul Mantz when Mantz opined to Goerner that a primary reason that AE was heard on 6210 was " 'It could have been they were too damn far away to be heard...6210 is usually good for just a couple of hundred miles' ". That would surely support TIGHAR'S theory she was enroute south towards NIKU.

Respectfully,
Ron Bright

From Ric

Last things first:

I don't know any way of checking Goerner's allegation about what Mantz told him. They're both dead and we've been unable to document other claims that Goerner made.

Commander Thompson's report (he was captain of the Itasca but his rank was Commander) to the commandant of the 14th Naval Distict is dated 29 July 1937 and he does say that he "Commenced laying heavy smoke at daylight." However, in his much more detailed report to the Commader of the Coast Guard's San Francisco Division (his own boss) he clearly gives the impression that, at the time of Earhart's presumed closest approach to Howland (between 8 and 9 a.m.) "Itasca was laying down smoke screen stretching for ten miles. Smoke remained concentrated and did not thin out greatly." (page 43 of Radio Transcripts Earhart Flight) That impression is almost certainly false and misleading.

This press release was sent by a reporter on Howland Island:

RDO USCG ITASCA CK 385 PRESS COLLECT 0300 HST 3RD BT; UNIPRESS HONOLULU PRESS COLLECT JULY 2; COPYWRIGHT [SIC] STORY UNDER SIGNATURE JAMES CHRISTIAN KAMAKAIWI;
WE WERE UP BEFORE DAYBREAK THIS MORNING THROUGHOUT THE NIGHT AMELIA WAS CLIPPING OFF THE MILES TO HOWLAND AND REPORTS SHOWED SHE WAS NEARING HOWLAND RAPIDLY WE WERE EXCITED AND I WAS PARTICULARLY EAGER BLACK HAD GIVEN ME THE HONOR OF WELCOMING MISS EARHART AS CHIEF RESIDENT OF THE ISLAND BOATS PUT OFF FROM ITASCA AT DAYLIGHT MEN WERE HURRYING TO POSITIONS AND AT SEVEN THIRTY HST WE WERE READY EVERYONE SEEMED TENSE AND SORT OF BREATHLESS WE WATCHED THE SKY HOPING TO PICK THE PLANE OUT AGAINST WHITE CUMULUS CLOUDS WHICH WERE ALL AROUND THE HORIZON THE SUN WAS HOT ON THE WHITE CORAL ITASCA WAS LETTING A BIG STREAM OF BLACK SMOKE OUT STREAMING LOW OVER THE WATER WITH THE TRADE WORD FROM THE SHIP AT SEVEN FORTYFIVE HST AMELIA ONE HUNDRED MILES AWAY WE WAITED NOT TALKING VERY MUCH BIG BOOBY BIRDS AND FRIGATES SOARING HIGH UP AND FAR AWAY LOOKED LIKE PLANES HOPES WERE RAISED SEVERAL TIMES BUT NO AMELIA WE WERE WAITNG NEAR THE WEST END OF THE EAST WEST RUNWAY ABOUT HALF MILE FROM THE GOVERNMENT HOUSE EIGHT THIRTY AND THE MINUTES DRAGGED THEN WORD WIGWAGGED FROM ITASCA AMELIAS SIGNALS ON DIRECTION FINDER SHOWED SHE WAS NORTHWEST OF THE ISLAND HAD SHE OVERSHOT TO THE NORTHWEST WAS A BIG BANK OF CLOUDS WHAT A GRAND BACKGROUND THAT WOULD MAKE WHY DOESNT SHE COME THE WIGWAG MAN WAS FLASHING TO US FROM THE GOVERNMENT HOUSE THE RECEIVER SHOUTED TO US AMELIA BELIEVED DOWN ALL SHORE PARTIES RETURN TO SHIP MY HEART STOPPED BEATING IT DIDNT SEEM REAL MEN WERE RUNNING TO THE HOUSE BOATS PUT OFF FROM ITASCA NO ONE WAS LAUGHING ORDERS WERE PASSED SHARPLY AND BEFORE WE REALIZED IT THE LOADED BOATS WERE BACK AT THE ITASCA EIGHT OF US COLONISTS WERE LEFT BEHIND WE WERE ALONE AGAIN ON THE ISLAND THE NOISE OF THE BIRDS SEEMED LOUDER WE WAITED NO VERIFICATION FROM THE ITASCA WHICH WAS LYING ABOUT HALF MILE OFF SHORE I COULDNT MAKE MYSELF BELIEVE AMELIA HAD MISSED US WE KEPT WATCHING THE SKY AT ELEVEN THIRTY SEVEN HST THE ITASCA STARTED OUT TO SEA TOWARDS THE NORTHWEST SOON SHE WAS DISAPPEARING OVER THE HORIZON WERE WAITING I HOPE WILL ALL MY HEART THEY FIND HER

Note that the reporter is referencing HST (Hawaiian Standard Time) which was an hour later than the time zone Itasca was using. This agrees with the ship's deck log that has the Itasca begin laying smoke at 06:14 local time. The reference to a DF bearing that put the plane northwest of Howland is interesting. The radio log makes it clear that no such bearing was taken. The ship's deck log also specifies that Itasca left its assigned station at Howland Island at 10:40 local time to begin searching to the northwest where Thompson had apparently convinced himself that the plane was down.

LTM,
Ric

Alan: You make some very good points, and all are essentially correct. I've no issues with what you said. But, if FJN had all of these tools at his disposal (and there is no evidence of same from other charts used by FJN for the AE flight), then why didn't he get to the island? You simply state that they just didn't see it within about 10 miles. I think that AE and FJN would have seen the high silhouette of the Itasca from at least that distance, as evidenced by the Colorado pilots seeing their ship from 35 nm distance. I think FJN did not have all of the tools he could have at his disposal, did the best job he could, but he also knew that he needed radio bearings for the last little piece of navigation and they simply didn't get it.

From Ric

As I think we've discussed before, what makes a ship visible from the air is not the high silhouette. You're looking down and there is no silhouette against the horizon. What you see is the white wake streaming out behind. Colorado was underway (or more correctly "underweigh"). Itasca was not. On the other hand, Itasca was painted white which should help some, while Colorado was presumably "battleship gray."

Anyway (anyweigh?), I think we're all arriving at the same conclusion. Noonan should have been able to get very close and there's no reason to think that he did not get very close, but it's also clear that he didn't get close enough.

LTM,
Ric

The 0718 position is also almost exactly on the great circle path from Lae to Howland: what a coincidence. It makes me very suspicious that this was a dead reckon position based upon speed over ground and/or elapsed time (some time in the past). I've checked all US and British maps from that era, and none show the Nukumanu Islands in sufficient detail to make a visual position report with the accuracy cited.

From Ric

Good point. In order to distinguish one island group from another you'd need to either be high enough to see the pattern of the entire archipelago with an unrestricted view (and Earhart specifically said she was at "8000 feet over cumulus clouds") or you'd have to have detailed planform maps of the individual islands (extremely unlikely).

Another Earhart myth bites the dust.

Alan C wrote:

>While going through this exercise it also dawned on me that everyone
>has pretty much made the loss a navigation issue.

and

>In our rush to explain why they missed Howland we have
>stripped Noonan of all his great experience and expertise and the tools of
>his trade.

Alan, well said -- I'm 100% with you on this one. The more I see, the harder it is to understand the loss from a navigation standpoint. Quite the opposite -- the departure time from Lae, the post dawn arrival time in the vicinity of Howland, and the daylight fix option could hardly have been more advantageous. Masterful timing (yes, and lucky) as far as it went. We've learned something about the equipment and challenges, but have not found any flaw that seems more likely than any other type of problem that could have done them in.

Looking at this is like trying to put together the plot of a full length movie by looking at a few individual frames. Earhart certainly could not be accused of burning up the airwaves with excess chatter! What little we have has long been pounced on and torn to shreds, and anyone looking at things today will probably start with a lot of inherited biases and assumptions.

Well, all that aside, it is time for me to return to being a "list lurker." My dear neglected wife and my poor dog were about to sign me up at the nearest support group for People who Post Too Much (maybe they even have an Earhartforum subgroup?). My teenagers were all too pleased at how easy it was to end any discussion of their whereabouts and activities by merely asking whether I had checked the forum email recently.

Take care,
Tom MM

From Ric

I do hope that you won't lurk too deeply. Your postings are most welcome and valued.

Looking at the language in the quotes from a military perspective (once upon a long time ago), the thing that now seems evident is the terminology "smoke screen".

Ron Bright wrote:

>Capt Thompson sent an official report to the Commandant 14th Naval
>District that on 2 July he began laying "heavy smoke screen at daylight."
>Period. Odd he uses "smoke screen" not "smoke plume".

Now I don't know about Cutters, but naval ships (ie: Destroyers) were equipped with smoke generators, which had nothing to do with the boilers or smokestacks --- these were for laying "defensive" smoke on the surface of the water. The smoke, which I seem to recall was oil-based, was intended to lie on the surface of the water to create a place for ships to hide. If in fact this is what did occur, there would be no thermal rise to the smoke, and until it settled it would just float along on the water. It also seems to me that there would be little likelihood that they would have fired up a lot of smoke from the boilers, if the Cutter was equipped with smoke generators.

>Itasca was laying down smoke screen stretching for ten miles. Smoke
>remained concentrated and did not thin out greatly."

This would be consistent with a smoke "screen", and is likewise consistent with the reporter's story.

>ITASCA WAS LETTING A BIG STREAM OF BLACK SMOKE OUT
>STREAMING LOW OVER THE WATER WITH THE TRADE

I presume the word "trade" is a reference to "trade wind". It would be helpful if he had included punctuation in his story. It seems to me that dark smoke lying close to the water on a bright morning, flying into the sun, would be really easy to miss.

The other thing that you may be able to clarify (since I'm a novice at the DF stuff they had) is the reference to the message from the ship to the island

>WORD WIGWAGGED FROM ITASCA AMELIAS SIGNALS ON DIRECTION
>FINDER SHOWED SHE WAS NORTHWEST OF THE ISLAND

Would the DF only show NW, or do you think they are "interpreting" the DF reading to be NW, when she might actually have been SE of Howland?

LTM,
jon 2266

From Ric

Interesting point on the DF. It may be that Itasca was able to get some rough bearing that was too shaky to include in later reports. If so, there may well have been a 180 degree ambiguity problem. Comments from the DF Delegation?

Not being a warship, I would be very surprised if Itasca was equipped with a "smoke generator." Here is a posting from last March by retired Destroyer-Escort captain Bob Brandenburg on the subject of smoke:

Black smoke makes sense in terms of trying to make the ship most visible to an observer. But I would be surprised if Itasca could make black smoke for an hour and a half. In a steam powered ship, black smoke is produced by reducing the amount of air in the fuel-air mix being pumped into the boiler fire box. This makes a lovely dense black smoke, but also deposits a coating of thick soot on the boiler tubes inside the fire box and hastens the time when the boiler must be shut down, opened up, and sailors have to climb in an scrape the soot of the tubes by hand, with wire brushes, etc. From your description of Itasca's size, I surmise she was medium endurance cutter, which means she had a single crew driven by two boilers. Although she would certainly steam on a single boiler while on station, to conserve fuel, the Captain would NEVER take a boiler off line for tube maintenance while at sea, because if the steaming boiler were to sustain a major casualty requiring it to be taken off line, the ship would be dead in the water. So, cleaning fire box tubes (known as cleaning firesides) was done in port, not at sea.

Fireside cleaning usually was done was done at intervals of 600 steaming hours. For a CG ship on extended station duty, that would interval would come up every 25 days. The on-station time could be stretched to 50 days by rotating boilers. But making heavy black smoke for a protracted period - - - more than 15 minutes or so - - - is inviting trouble in the form of a thin-lipped tube rupture (caused by uneven heating of the tube surface due to the soot accumulation) which results in water and steam spewing into the fire box, dousing the fire and, worse, causing the firebrick lining of the fire box the crack from chill shock and crumble into a pile of rubble in the middle of the firebox. Not a pleasant event, one which requires major and expensive shipyard repairs, and one which skippers of steam ships avoided like the plague.

> I begin to wonder if Itasca was making any smoke when it might
>have done some good.

Bottom line, I agree that Itasca probably was not making smoke when AE needed it. That can be verified from the deck log.

But even it Itasca was making smoke, the wind-driven downsweep of the smoke, which I suspected and your photo confirms, would have caused the smoke plume to widen rapidly, and become quite thin in the vertical plane. That would drastically reduce its visible contrast with respect to the sea surface, thus rendering it virtually invisible except at close range. Smoke is really only useful as a visual detection aid when the wind speed is less than about 5 knots, so there will be a significant vertical plume. Then it is readily visible in contrast to the sky luminance.

The fact that Itasca was painted white and was relatively small would make her blend in with the whitecaps, and maker her nearly invisible beyond a few miles. The problem was compounded by the fact that the wind blows almost continually, albeit at varying speeds, in that neighborhood, creating a substantial fetch that produces well-developed swells. So, when the swell has white caps, the result is what looks like many parallel lines of white on the surface. It's really tough for a pilot to see s small white ship in that background. If Itasca had been steaming at 10 knots or so, her Kelvin wake would have been easy to spot. But a drifting target is nearly invisible.

So, there you are. Unless FN's navigation was perfect and they overflew the island, they had virtually no chance of seeing either it or the Itasca.

Bob

TRIGONOMETRY: Can somebody out there dredge up their high school trigonometry and figure out how tall an object would have to be to be seen at sea from a distance of 20 miles and at 40 miles at sea level and at altitudes of 1,000, 5,000, and 10,000 feet? I know you have to figure in the curvature on the earth on this one, or else our geometry experts could weigh in on the problem.

SMOKE: Also, as Ric has pointed out there are some CYA statements in Commander Thompson's report on the Earhart flight. There are also a couple inconsistencies that even a novice can pick out.

"[L]aying down smoke screen . . .": How can the Itasca lay down a smoke screen if she is not moving?

The general idea of a smoke screen, or in this case a beacon, is to use the ship's forward movement to disperse a sufficient volume of smoke to be seen from afar by other ships and aircraft. The effectiveness of a smoke screen depends on the wind conditions, the ship's heading and the amount of smoke it can make. But most importantly it depends upon the ship to be moving.

In a no-wind situation, you can make smoke sitting still but I suspect all you'd do is foul the deck with soot, as the smoke settles on the stationary ship.

And even if you had a wind it would foolish to rely on the wind alone to make it an effective "smoke screen." The idea of a moving ship is to put the smoke screen where you want it.

" . . . stretching for 10 miles": Hm-m-m, this too sounds self-serving. Here the commander has quantified his actions. He is asking us to believe that the Itasca was creating a smoke plume thick enough (sufficient particulate matter in suspension) and with enough heat energy (BTUs) to stay aloft and stretch out 10 miles -- all while remaining stationary. That would be quite a breeze to accomplish that.

Standing on the bridge and looking downwind, I can see how the commander may believe he was creating a smoke plume visible from 10 miles. But how could he tell it was "stretching for 10 miles." Even if the plume did drift for 10 miles I am not convinced that it was still compact enough to create a visible beacon. At 10 miles, I suspect it to have diffused considerably to the point it blended in with the ambient haze, clouds, etc.

Just some random thoughts.

LTM, who never got past algebra
Dennis O. McGee #0149CE

Dennis asked:

>TRIGONOMETRY: ......how tall [would] an object......have to be to be seen
>at sea from a distance of 20 miles and at 40 miles at sea level and at
>altitudes of 1,000, 5,000, and 10,000 feet? I know you have to figure in
>the curvature on the earth on this one, or else our geometry experts could
>weigh in on the problem.

Not only curvature, but refraction must be figured in.

Geometrically, the nautical mile distance to the sea horizon is about 1.06 times the square root of the altitude. (1.07 times the root if at or above 10,000 MSL). That accounts for curvature. But because of refraction, the prefactor is increased to about 1.14 or 1.15.

At Sea Level, to see an object 20 miles away the formula indicates that the object must be at least 302 feet high. To see an object 40 miles away it must be 1209 feet high.

Flying at 1000 feet, a surface object 20 miles away might be visible since the visible horizon is 36 miles away. To be visible 40 miles away it would need to be at least 100 feet high.

Flying at 5000 MSL the visible horizon is about 81 miles away; at 10,000 MSL it is 115 miles away. From those altitudes, objects at 20 and 40 miles don't need to be elevated to be brought into view.

Mark

Let's see if a layman has understood some of this navigation stuff: I'm going for a quick flight from Lae to Howland Island...

I've been flying all night and I have a DR position that puts me somewhere near Howland, on a line drawn from Lae to Howland --- but I don't know exactly where I am on the line because of variations on aircraft speed and track due to winds aloft. I know the distance I had to cover, and I know the speed I should be flying and I have the direction drawn as a line on my chart (map), so I know how long the flight should be.

I think the first thing I should be able to get a good fix on is the bottom of the sun just clearing the horizon in the morning, and I can measure an angle called the azimuth with my trusty Pioneer Bubble Octant (Model A5). (Details of how this is done and what the azimuth is have been covered in the forum).

During the night I picked a time that should put me somewhere short of Howland. Looking at my tables I saw that at that time, the sun's azimuth (that angle again) would be 67 degrees. I drew a line at right angles to the 67 degrees and that line cut across the line that shows our track on the chart.

I will wait until the bottom of the sun just clears the horizon and note the exact time.

Now this "exact time" is very important as it tells me pretty much where along the line on the chart my Line Of Position Falls.

Once I get the sight I can then move the LOP I drew on the chart to Howland (along the line I drew that shows my direction) and see that there as X miles separating the two lines. That is how far I should be from Howland. I will do a simple time and distance calculation and I will know I have X minutes to fly to reach the line that I have moved ahead and now lies through Howland.

So Far - So Good...

It is just light in the east, there are clouds on the Horizon. I check the time, yes, according to the tables it is time to "shoot the sun", but I can't get the exact moment that the sun clears the horizon because the clouds are in the way.

I can wait a while, because the sun will be at an azimuth of 67 degrees for a little over 20 minutes and in that time the bottom of the sun's disk will move to about 4deg above the horizon. However in that 20 minutes, the line on the map across my path will move something more than 300 miles whilst at the same time my aeroplane will move around 30 miles.

Which kind of screws up my calculations. Now I know by Dead Reckoning that I am close to Howland, but I have no way of knowing if I am almost there, just past it, or slightly North(ish) or South(ish) and if I do get another shot at the sun, I can't get any more proper fixes on it for about an hour.

So we fly around for a while and look for the Island and a ship. While this is going on, I work out how much fuel we have left, and how long we can keep flying. Looking at where we should be on the chart, I measure a line from there to the end of the distance we can fly on the fuel we have left and see if any "easy to see" land falls within that distance.

Which is the point of my other post. Until I ran the figures I thought the sun was at 67deg azimuth to a given (stationary) position only when it touched the horizon. The calculations show that it was there for a long time (until it reached a bit over 4deg declination. then it returned to 66deg azimuth and so on.

The moment clouds are on the horizon there is a major problem for FN, and the reporter on Howland reported in his report that there were clouds "all around the horizon". If the reporter's report was accurate (I wonder how many are) then Fred Noonan possibly didn't get that azimuth shot as the sun just cleared the horizon.

Which kind of means falling back on our trusty Direction Finder...... And finding some land to bug out to fast as soon as our fuel gets down to reserves....

RossD

From Ric

Okay, that's pretty much what we've been saying except that I would think that Noonan would have his "Plan B" worked out in advance and running southeastward on the LOP is not so much an admission of defeat but a sure-fire way to find land of some kind, and you still hope it will be Howland.

> I've been flying all night and I have a DR position that puts me somewhere
> near Howland, on a line drawn from Lae to Howland

Ross, that was a pretty fair understanding of all the gobbledy gook we have presented to you.

A couple of comments, however:

We are still slighting Noonan a bit if we think he flew all night and has only a DR position to work with. That may well be true but we have no concrete evidence he didn't have good solid fixes along the way. I sort of think if I had been flying from Lae to Howland and couldn't get ANY star shots because of weather that would have been the time to put down somewhere until I HAD good weather. That might have even meant turning back but better than blindly flying into oblivion.

Second point is the sun shot on sunrise. Right as the sun peeps over the horizon you can check the azimuth but timing it to get a position is not accurate enough to bother with. The Naval Observatory says that time could be off a couple of minutes and that's not good enough. Until the sun reaches altitude the shot could be unreliable. I suspect because of the timing between sunrise and 8:44 local Noonan DID shoot the sun at a very low altitude. I say that with fair confidence because the sun WAS at very low altitude.

Having said that I think you're on the money as to basically what transpired. Noonan did his job, got the plane somewhere close (whatever that means) and for whatever reason could not see the island and went elsewhere.

In the past notes we have pretty much agreed they would not have gone back to the Gilberts. I think the reason we have said that is that (1) the islands are widely scattered increasing the chance of missing them and (2) Noonan would have again been constrained by sun LOPs with little north/south information. But actually we need to consider that as he might of had as much info going back to the Gilberts as he did coming into Howland in the way of other planetary bodies from which to obtain a fix. Soon he would lose the moon, however, and I have not checked to see how long a planet hung around but a bright sunny day would have made planets less than easy.

Another reason not to head for the Gilberts might have been the distance being greater than down to the Phoenix group. That brings us back to the fuel question. I might have thought hard about finding a usable runway as a first choice alternate if I had the luxury of fuel and I would suppose their planned alternate or alternates would have included a runway someplace but any ground is better than none.

Alan
#2329

From Ric

In 1937 runways were in rather short supply around the Pacific. There was Lae and there was Howland. That's it. Nothing in the Solomons, nothing at Nauru, nothing in the Gilberts, nothin' nowhere.

>From Alan Caldwell
>
> In 1937 runways were in rather short supply around the Pacific. There was
> Lae and there was Howland. That's it. Nothing in the Solomons, nothing at
> Nauru, nothing in the Gilberts, nothin' nowhere.
>
> Wouldn't you have loved listening to them discuss their alternate?

Carrington in his book asserts that AE and FN did consider the Phoenix islands as their alternate for the original East to West leg from Hawaii to Howland. The impression he gives is that this information came from Paul Mantz, who was still involved with planning during the first attemept.

The point is that they may have had the Phoenix islands in mind as an alternate long before they even started the second attempt, and almost certainly not some sort of last ditch, spur of the moment decision at the time they couldn't find Howland.

Didn't we try to figure out what alternates were available for the first attempt should they not be able to find Howland based upon expected range etc. I cannot remember what we decided, are the Phoneix Isles even a possibility as an alternate for the East to West leg? Any other possibilities?

LTM ( who cannot remember everything.)
Andrew McKenna