In re Ric's analysis of LOP limits:

>Again, assuming that you are someplace on the line, what is the greatest
>distance you can be from a dot? Three and a half inches (350 nm) if you are
>at the extreme left end of the line.

If you don't know where you are on the line, you could be farther than three and a half inches (350 nm) to the left of the center dot. If this is possible, I wouldn't waste any fuel heading left (northwest); I would head right (southeast), since I have a better chance of seeing something along the line in that direction, and, possibly, no chance of seeing something in the other direction.

David Evans Katz

The significance of the 350 nm is that it is the distance between Howland (the desired destination) and Gardner (the farthest practical alternate on the line). If you only explore southeastward on the line you abandon the possibility that Howland is just over the horizon to the northwest, and you'd much rather reach Howland than anywhere else.

Next time the subject comes up, it might be helpful for some to know that although an lop for any purpose incl. part of a fix, while drawn on the map as a fine line, should be thought of as a 'swath' having a dimension of width equal to the approx. potential errors. When the sight was steady and well done it should be only a mile or two in width. A 'shaky' sight [turbulence, drifting clouds, etc] might be 10 or more miles wide. Regardless of initial accuracy it has the potential to 'grow wider' when advanced, depending on how accurately the ground speed was forecast, [as you mentioned] and miles of the advance.

One can expect greater error using a hand held insrtrument thru a slanted windshield; compared to having it hanging from a hook in an astrodome, in which the optical errors of the dome medium have been accurately determined. Clearly, FN was operating in less than ideal conditions.

You probably know of the many errors in getting the line properly drawn on a map. Beginning with the rounded values in the tables, the required math, the exact time, instrument & optical error incl. 'thru a windshield or any medium', and sometimes, parallax.

Averaging the sight may be the largest; the body can never be held in the cross hairs for more than a couple seconds, even in still air. And there are a few more. Fortunately, the errors are seldom all in the same direction. Usually, but not always, some average out others.

RC

Ric said...

>Fortunately, there's little reason to think that anything is buried in that
>sandbar, according to a PhD reef biologist at the University of Hawaii who
>has considered the situation. Material washed through the passage should be
>on the lagoon bottom at the base of the "talus slope".

So, it's the talus slope that is probably of most interest. I take it the pump is not adequate to be of any use examining the talus slope either. And you wouldn't want to move enough material to do any good anyway. Whatever a diver can see is about all that is left.

It would be interesting to know whether or not that sandbar has been growing and continually covering whatever might have been at the base of that talus slope at one time in the past. If that has been happening, there's little hope of finding anything that washed into the passage, be it pieces of the Norwich City or Earhart's Electra.

Correction: He's not with U. of H. Here's exactly what Jim Maragos, coral reef biologist with the U.S. Fish and Wildlife Service, Pacific Islands Ecoregion said:

The most likely places to search for wreckage would be, in descending order of likelihood:

a) base of lagoon talus slope downdrift of, or opposite the landing site,
b) reef flat at landing site
c) deep ocean reef slope beyond the reef terrace
d) reef terrace just downslope of a water landing site
e) embedded in sand bar or island deposits

In contrast, the pass or lagoon bars inside the pass would be very unlikely to yield wreckage, unless the plane landed in the pass itself.

Tom, I guess I wasn't very clear but I DID recognize the 20% figure was not something you were suggesting. What I was trying to say was the actual figure could only be 0% or 100%. In 1976 my son was diagnosed with cancer that had a 12% survival rate. That was a totally useless figure. My son would survive or he wouldn't. It also represented a statistical average which again is of no value to me. To someone but not to me. My son was lucky and he DID survive but at no time were decisions based on the 12% probability.

I agree with you that some analysis might be in order but only if we know the input data and the analysis provides any useful information.

Your example of a target 2,000 miles away DOES illustrate your point but that was not the case at hand. FN's target was but a fraction of that away. I think we need to use actual facts if any analysis is to be meaningful. Therein lies the rub. We don't have enough information to analyze.

I think in a previous post I explained how a fairly accurate positioning of the LOP would be obtained. Now since we don't know the weather we don't know what fixes FN got if any. That means he either did and got an accurate LOP or he didn't and DRd to someplace he hoped Howland was. We will never know the answer to that unless we find the Electra and FN's maps.

If Noonan was able to get a couple of sun shots then he would also have been able to shoot the moon and Venus barring cloud cover in their directions. If that was the case he most likely had a good LOP but possibly 5 miles from where Howland actually was situated. And again we don't know if he had the corrected coordinates or not. The moon and/or Venus would have given him a fairly accurate position so it is doubtful he ran north and south on the LOP very far. More likely he ran it in a modified search pattern, varying the "LOP" a few miles in either direction so as not to lose his position. There was time between the "We must be on you" call and the "running north and south" call to do a minimum search before striking out.

As to their options I think there was no choice other than to find Howland or head for the Phoenix group. Ditching in the ocean was not likely a choice not were the unreachable Marshalls. The Gilberts made little sense even considering the moon and Venus being available to help with the navigation as they were only available in the early hours. By the time they could get to the Gilberts FN would have only had the sun and a very unreliable position. Choosing the Phoenix group made sense in that the sun provided a course and there were several islands in a loose group.

I don't think it would take any kind of deep analysis to make that choice and I could bet FN never thought about odds or percentages. But let's suppose he did or we wanted to assess those odds how would one do that and to what end?

First off, I'd like to say that your posting of a few days back regarding the casting of a wide net (wherein you noted crashed/sank, jap capture, and all the Phoenix islands) was an excellent summation of TIGHAR's Earhart Project to date. In essence, you did cast a wide net, and Niku is all you caught in it. Perhaps you could arrange things so that that post appears on the screen of anyone submitting a posting to the forum, before the listserver will accept the post, just as a friendly reminder of why we're here and what we're trying to do.

My question is for the Noonan Project team and the Celestial Choir. If I'm following the threads correctly, the flights where FN's nav practices are being examined are the Oakland/Honolulu success, and the Lae/Howland failure. For Oakland/Honolulu, Fred used 2 body fixes, and had radio bearing assistance for final approach, correct? Is it then assumed that expecting radio bearing assistance on the final approach to Howland, that he used 2 body fixes for Lae/Howland? If so, given the greater degree of difficulty of the flight (Howland after 2500 miles being a more difficult target than the Hawaiian Islands after 1500miles) is that assumption a reasonable one. I'm not nitpicking--just curious as to your reasoning. You guys have done the research, and I'll accept your conclusions.

Finally, for the presumed Howland vicinity/Gardner flight, Fred would (obviously) not be expecting radio bearing assistance for the final approach. Since this flight was during daylight, what sort of celestial nav./DR/etc. did FN have available? Is there any record of FN using more precise techniques, i.e. 3 body fixes, on flights where radio bearing assistance was not anticipated for final approach? I'm guessing that one cannot get a 3 body celestial fix during daylight, correct? Was there any technique available to FN to back up the LOP run to Gardner that might give him a greater certainty of success? Granted of course that if something was available, whether or not he used it would be pure speculation.

Sorry for all the questions, and for any gross oversimplification errors I may have committed above. Just trying to get a better handle on all the nav. stuff on the forum.

Dave Porter, 2288 (whose nav. skills are limited to compass, map, protractor, terrain association, and "stoutish walking shoes")

A couple of points:

Oakland to Hawaii is more like 2100 miles.

Noonan most definitely WAS expecting radio bearing assistance for the final approach to Howland. That was the whole point in Itasca being there.

Dave Porter raises excellent questions. Let me try to answer them briefly.

As Ric has already pointed out, the Oakland-to-Honolulu and Lae-Howland distances were comparable. And Noonan definitely was assuming that there would be a radio bearing to guide him to Howland at the end of the trip.

The Oakland -- Honolulu flight is the only documented example we have of Noonan's navigation practices. What we see there is that he never used more than two celestial bodies in a fix (a two-body celestial fix taken with a hand-held bubble octant can be in error by 20 miles or more), he allowed intervals of 2 to 3 hours between fixes, and that he assumed that he would get a radio bearing to steer at the end of the flight. Absent evidence to the contrary, we assume that this was typical of Noonan's navigation.

If Noonan didn't know where he was at the time he expected to get the radio bearing from the Itasca, it would not be surprising that the best he could do was to get a sun line and say that he was somewhere on the 357/157 LOP.

We are not aware of any instances in which Noonan used 3-body fixes, whether or not radio bearing assistance.

Dave is correct that a three-body fix in daylight is not feasible. Noonan did have the sun and moon available on the fateful morning.

As for what Noonan could have done to back up the LOP, he could have prepared for the worst case , i.e. that the expected radio bearing from the Itasca would not be available, and take very frequent fixes to reduce the cumulative position error to a manageable level.

Hope this helps, Dave.

Bob Brandenburg, #2286

We also have the South Atlantic crossing, as documented on the chart he actually used (now at Purdeue). Another long, over-water flight but with no expectation of a radio bearing at the end. It also has the advantage of there being no question as to who is doing the navigating. I don't have a copy of the entire chart. Randy, do you?

> Fortunately, the errors are seldom all in the same direction. Usually, but
> not always, some average out others.

Bob, that was an excellent posting on the problems of celestial as it pertains to our subject. I hope everyone can see that even with everything going for Noonan his navigation to Howland might at best put him in the general vicinity of where Howland actually was or where it was erroneously plotted. It then depended on a visual sighting either by the crew or those on the ground or a DF steer. Clearly none of that was forthcoming.

The more this subject is looked at I think it becomes clearer that hitting Howland was a far more difficult task in 1937 than they realized. It also is obvious that losing the belly antenna, not assuring DF was working and not establishing two way communication combined into a fatal scenario. From take off there were many opportunities to resolve those difficulties but no attempt was made to do so.

It is possible that the lower antenna snapping off made no audible sound and that no one on the ground watching the take off saw it. I have my doubts. AE wasn't much for radio communications and there probably wasn't all that much in the way of position reporting in 1937. I am only guessing at that. Also, in spite of never being able to talk to anyone she was still stubbornly broadcasting as though she couldn't quite get it through her head that she had no two way capability. The DF never worked on their trip so I can only assume FN was counting on a steer from the Itasca. It's hard to imagine planning a more unsuccessful flight.

Alan
#2329

Dave Bush wrote:

> Finally, for the presumed Howland vicinity/Gardner flight, Fred would
> (obviously) not be expecting radio bearing assistance for the final
> approach.
>
> Since this flight was during daylight, what sort of celestial nav./DR/etc.
> did FN have available? Is there any record of FN using more precise
> techniques, i.e. 3 body fixes, on flights where radio bearing assistance was
> not anticipated for final approach? I'm guessing that one cannot get a 3
> body celestial fix during daylight, correct? Was there any technique
> available to FN to back up the LOP run to Gardner that might give him a
> greater certainty of success? Granted of course that if something was
> available, whether or not he used it would be pure speculation.

Dave, Ric answered your questions about FN expecting a DF at Howland. I think it is pretty clear he did. More than that it appears he NEEDED it.

As to what Noonan had available upon reaching the vicinity of Howland he had the sun, moon and Venus clouds permitting. They would not have given him more than a two body fix due to their positions but he could have used all three to refine his position. If he had visibility to shoot sun shots he would have also had an east/west wind component and a ground speed. It is possible to get a three body fix in the daylight depending on the location of the bodies. Ideally we looked for a fairly equal triangulation if possible.

I don't know what FN's navigation was like but some here on the forum are better qualified to answer that than I am.

On the flight down to Niku FN had the same capability but no more. He would have repeatedly shot the sun and whatever else was up to navigate to the Phoenix group. He should have had a course line from the sun and still an east/west wind component. If the planet Venus was visible he could have obtained a speed line and a two body fix. I would have to go into the charts to see if he had more than the sun on the way to Niku.

Something I have always fretted about was the visibility out of the airplane to shoot celestial. I don't know what the limitations were. Maybe someone here can answer that. I would suggest everyone keep in mind that wherever the celestial body was directionally the airplane could be turned to that was not a limitation - only an aggravation. The real question is how high an angle could Fred shoot. I have recently been advised by a long time military navigator that a shot could be taken in a slight but constant climb and certainly a slight roll with no heading change could expand the available angle. Only the sextant needs to be level. I'm not sure AE would have had sufficient skill to do either but it remains a possibility. If anyone wants to refute or support that concept I would be glad to hear from them.

Maybe I'm just dense, but assuming I'm somewhere on that line, but not knowing exactly where, how would I determine whether I'm not already at the extreme left (or NW) end of the line, when my chronometers inform me we should be intersecting with Howland (the dot in the middle of the line)?

According to your explanation, it would seem I would have to have some reasonable knowledge of just how close to Howland I am on the LOP, in order to determine which way I should turn & for just how long I could maintain such a course, because if in fact I'm already at the NW extremity of the 7" LOP line you've drawn, I'm already 350nm NW of Howland & turning NW (which we must presume they did, given AE's ...'running north & south'... message) for any extended period could seriously affect my ability to ultimately reach Gardner at the opposite end of the LOP, if after turning SE, I happen to miss Howland/Baker a second time.

Is it possible that as the Electra approached Howland & assuming that FN had not calculated a deliberate off-set, either NW or SE of Howland, could he have determined that they were on either the NW or SE side of Howland simply by the position of the aircraft, with reference to the rising sun, which if I recall correctly, should have been directly in front of them, if AE had accurately maintained their originally charted course during the night time hours of the flight... of course not considering any significant wind drift during the night... or am I again oversimplifying ?

Frankly, I just find it hard to believe that FN could have been off course 350nm to the NW , if the much later reports from residents of Tabiteuea, in the Gilberts, about hearing an aircraft over their island during the night, is true.

Don Neumann

No, Noonan could not tell which side of Howland he was on based upon the position of the sun because its progress across the sky was not enough different from where it had been at sunrise to give him a meaningful "cut."

I too doubt that Noonan was 350 miles off when he hit the LOP but the point is that he couldn't know that for sure. (Even if the Tabituaea report is accurate, it was pitch dark when they crossed that area and they had no way of knowing that somebody heard them.) Not knowing where he was on the line, he did know that heading southeastward on the line while he still had enough fuel to go 350 miles was the best way to guarantee that he would reach some kind of land.

I have copies of both the S. Atlantic crossing and the Oakland-Honolulu crossings, but they are too large to conveniently copy or scan electronically.

Parenthetically, I should mention that I've not done a rigorous analysis of the Oakland-Honolulu map writings to determine whether it was FJN or Manning who made the markings. They do look very similar to later writings that we know came from FJN.

FJN did attempt a couple of 3-body fixes, but they were not satisfactory. The S. Atlantic crossing has but one fix or observation in the middle of the ocean that I cannot explain how FJN came to that particular point, as the "planets" don't line up.

They were also reportedly in solid cloud for almost the whole trip.

Having flown in the Lockheed 10A (which has the same cockpit windows as AE's Lockheed 10E) I can say that the view from the cockpit is good and far better than from a DC-3. The windscreen is high and wide and so are the both side windows (which can be opened in flight). Sitting in the right hand cockpit seat FN had excellent view forward and to the right and using his sextant should not have been difficult. As the sun climbed AE could have lifted the nose of the aircraft to facilitate FN's work (which would have resulted in an increase of altitude).

When looking at pictures of AE's Lockheed10E one can see that it also had two large observation windows in the rear fuselage, one on each side. If Fred was not sitting up front he would have been able to make fixes from his navigator's position in the rear. I'm not sure what angle the fuselage windows allow but I agree that it would have a simple matter for AE to tilt the aircraft a bit and allow FN to have a better view of any celestial body available from one of the windows. Remember that the cockpit windows could be opened for a greater obvservation angle. I believe the starboard fuselage window was in fact an emergency exit and the Electra not being pressurised, could be opened in flight if this was really necessary.

On the standard Lockheed 10 the emergency hatch was as you describe but that was not the case in Earhart's machine. The large nonstandard window installed in the lavatory area of NR16020 prior to the first world flight attempt and subsequently skinned-over before the departure from Miami on the second attempt appears to be the same dimensions as the standard emergency exit. I wonder if the skinned-over window was, in fact, a hatch that could be removed in flight for taking unobstructed celestial sightings. Hmmm.

> She knew that she was heard on 3105 because the
> Itasca transmitted the letter A on 7500 at her

Gosh, here I am wrong again. I knew that but it had faded off into oblivion. Doesn't that make it even more aggravating? If she had ANY clue someone heard her transmission on 3105 it seems she would have returned to 3105 and tried again when her "day" frequency produced no results.

Alan
#2329

Well, she DID return to 3105 at 19:30 GMT to say that she couldn't get a minimum. She may have tried again on 3105 at the next scheduled transmission time (19:45) but Itasca was blocking the frequency so they could not have heard her. She tried again at 20:13 to say that she was running on the 157/337 line and that she'd repeat the message on 6210.

Here's a thought. What if she had transmitted her earlier messages on 3105 and repeared them on 6210 (without mentioning that she was doing that)? It wasn't hard to change frequencies -- just flip a switch. If she did that then she would not have been sure which frequency Itasca had heard her on. With no subsequent luck on 3105 it would be reasonable to assume that 6210 was a better choice.

Ric wrote:

> ... What if she had transmitted her earlier messages on
> 3105 and repeared them on 6210 (without mentioning that she was doing
> that)? It wasn't hard to change frequencies --- just flip a switch. If
> she did that then she would not have been sure which frequency Itasca had
> heard her on. ...

Well, if that's what she did, then she shouldn't have stuck with one frequency. The best thing would be to keep broadcasting on both, just in case.

Itasca needed to get through to AE to deliver two messages:

1) You haven't given us a signal long enough for us to get a bearing on you.
2) Here is the direction you need to fly to find us.

Since neither she nor Fred was up to speed with CW, and since Itasca's voice transmissions seem never to have reached her, and since Itasca had no voice capability on 7500, I don't see how these two messages could have gotten through to her. The failure to master CW is one link in the chain of poor decisions that led to the fatalities. Both DF and radio telephony were "modern" techniques. Having less sophisticated (but more demanding) older skills available might have saved them.

Or, as other people have put it, they failed to have in place a reliable Plan B.

Marty #2359