The Electra Project (TEP) is building a digital Lockheed Model 10-E Electra Special, that flies in the X-Plane flight simulator program. We are using X-Plane for THE ELECTRA PROJECT because the program uses real flight characteristics, and the physics of flight to set up its digital models. X-Plane cannot model everything on the L10E “Special,” but the TEP model is as close as possible in its aeronautical design to the 1936 aircraft.
There is data available to build a digital model of the Electra, however, there are few facts to find the missing Earhart plane. The following narrative is based on the idea of using a flight simulator, known facts, and a number of previous observations, to learn more about flying the Electra. TEP chose the last Earhart flight to model, but the plane can also fly anywhere – even into the 21st Century.
This file is a brief compilation of data, and observations, recorded after test flights with the simulator. (Note: We have been flying again with the simulator, and The Electra Flies have been updated to reflect this activity. 3.25.2015)
Definitions for terms used here:
GCT: Same as GMT for our purposes. SoG: Speed over Ground. LOP: Line Of Position.
Usually, at the scene of an accident, there are clues about what happened. Yet in the case of the Earhart & Noonan disappearance the ocean has the scene under 16,000 feet of water. Beginning the effort at takeoff TEP has looked at their last flight from inside the simulator Electra cockpit.
Eric Chater, general manager of Guinea Airways, witnessed the Lae takeoff and documents most of what we know about their departure in his 1937 report. His narrative was in the form of a letter to M.E. Griffin, at Placer Management, in San Francisco. Griffin, in a radio message, after the Electra went missing, asked for details from Lae.
TEP Electra's takeoff weight from Lae is approximately 15,500 lbs.This weight includes 6767 lbs. of fuel (1100 Gal US), and a payload weight of approximately 860 lbs.This makes Earhart and Noonan’s departure weight a close call; they are overloaded!
There is a short film of this last takeoff. Earhart does not get airborne until her aircraft tires hit the crown of a road that ran along the edge of the coastline at the end of the field. She literally bounces into the air. The film also shows that she did not use flaps for takeoff. The Electra’s flaps were “split flaps,” and useful as air brakes to slow down the plane for landings – not for lift at takeoff.
Why did the Electra fly south from Lae?
Weather reports from the Navy in Hawaii place large clouds and thunderstorms 200 - 300 east of Lae at takeoff. This weather may have forced their flight south around these storms. Four hours after takeoff Earhart is reporting they have climbed to 10,000 feet.TEP’s Electra model shows this climb increases the Electra’s fuel burn rate by about 24% during the maneuver.
Most historians have discounted the southern waypoint as being in error; that the radio operator on duty in Lae may not have heard her radio report correctly. But when TEP plots the course over Bougainville, from the southern waypoint of 7.33S x 150.7E, and then to Nukumanu Atoll, the position appears OK. The time of the radio call, 0418 GCT, may not have been when they passed this point. We do not know the reason why, but the time is probably connected to her radio schedule (18 minutes past the hour), and not the position.
Later, passing Nukumanu Atoll, Earhart’s radio position of 4.33S x 159.7E, at 0718 GCT, the time of her scheduled radio call, do not (again) necessarily go together. Our Electra model, using an indicated air speed of 140kts, with easterly winds of 23 knots, gets her their earlier. We have learned, since test flying our Electra, the position coordinates were undoubtedly pre-computed by Noonan before takeoff from Lae, but that the actual chart position Noonan had for the atoll, was incorrect.
Up until, and including Nukumanu Atoll our model’s flight, and & position reports, coincide closely on all our test flights.
BACK TO TOP
The importance of observing waypoints along the flight should not be underestimated. Up until Nukumanu Atoll, Noonan is accurate with his navigation. It has been daylight, and not overcast. After leaving daylight at Nukumanu, all possible future waypoints are in darkness. If as reported, they flew in clouds, overcast (and perhaps even undercast), during the nighttime hours, accurate navigation was less likely.
TEP flies east from Nukumanu Atoll toward Howland Island almost solely by dead reckoning beginning sometime after midnight local time (1030 GCT) by enabling clouds in our flight simulation. They do apparently see the lights of the SS Myrtlebank south of Nauru. Accordingly we have moved our track slightly to the north. But it’s a guess.
None of the islands along the way, the Gilberts, had any significant nighttime lighting. Earhart’s radio call of 1415 GCT reports they were flying in clouds, and overcast during the last part of the route. We simulated similar conditions, but could not see much of the simulated overhead sky as we flew. Nor could we see the ocean surface to gauge drift or speed.
TEP believes the potential lack of celestial fixes, and observation of ground waypoints during the latter part of their night flight, has been under estimated.
At approximately 200 miles out, and at 1744 GCT, Earhart calls the USCG Itasca and asks for a DF bearing. It is not, however, her usual radio schedule. This request, however, would be Noonan’s standard PanAm DF procedure. But Earhart did not know how to use the radio direction finding equipment. No bearings were ever obtained.
No one should tie Earhart’s radio schedule times to aircraft positions. It’s tempting, but not useful. At approximately 200 miles out, and at 1744 GCT, Earhart calls the USCG Itasca and asks for a DF bearing. The timing of this call is not on her usual radio schedule.
Again, 1815 GCT Earhart calls (this time on her radio schedule time), and radios they are approximately 100 miles out. She asks for a radio DF bearing from the Coast Guard. She does not get a bearing. These two calls, with conflicting times, highlight the issue: You cannot fly this distance in the time between calls. We should not use these times to confirm navigation. There are probably only two correct times during the entire flight. The beginning at 0000 GCT, and when Earhart says they have arrived, at 1912 GCT
At this point, sitting in the cockpit of our digital Electral, TEP believes they are saying,
“We are lost. Take a fix on us so we know where we are.”
Without the aid of a RDF fix, standard navigation procedure of the day, would then have Noonan telling Earhart to turn north (or south) to fly an offset course. He would have taken at least one sun line observation at sunrise, and perhaps shot a second, to create a forward line of position. We cannot know how accurate those observations were.
Around 1900 GCT they may have turned onto that forward line of position believing it ran through Howland Island. At 1912 GCT Earhart announces that they must be on the Itasca’s position at Howland, and they are low on gas.
Out of 7 test flights in our digital Electra, two flights passed Howland Island to the east by approximately 25 miles, at 1900 GCT. Five flights ended up west of Howland.
What caused our TEP model to fly past Howland Island? Earhart reports an indicated airspeed of 140 kts. TEP calculates an average speed over ground of 113-117 kts. Change headwinds, and their SoG changes. We do not know what happened during the nighttime portion of their flight. But we do know that by changing the headwinds, with very small numbers, we could move the simulator position either east or west.
TEP suggests there is about a 75/25 chance of either arriving west, or arriving east, of Howland. Until now every search operation has gone west and found nothing; no one has looked east. For now, there is no way to know if this is important.
Beginning at 1912 GCT, and on to 2013 GCT, the Coast Guard radio operators reported Earhart’s radio signals at almost the same signal strength during each call. How is it possible a plane can fly for 1 hour and each radio call get about the same reception signal strength evaluation?
TEP believes there are two ways this might be possible: The first has the transmitter flying in circles
around the receiver. Since Earhart did not know where they were how could they fly in a circle
around an unknown point?
The second possibility we call the Pythagoras Theorem solution. Put the transmitter at approximately 30 to 70 miles either west or east of the receiver at the beginning of the experiment.
Fly the transmitter 30 miles to the north on the LOP. After the transmitter reaches 30nm miles to the north, turn around and fly the reciprocal to the south for 60 miles. The transmitter will have then flown over 90 miles on the LOP. You create an almost right triangle — with the distance between the transmitter and receiver the hypotenuse of that triangle.
But the actual distance, the hypotenuse, from the transmitter to the receiver, will change very little over time despite the fact that you flew 90 miles!
TEP concludes Earhart and Noonan most likely were NOT due north, or due south, of Howland when they ran out of gas, but somewhere either more or less east or west.
TEP believes the downed plane may be either west or east of Howland at some distance –
perhaps between 30 and 60 miles.
BACK TO TOP
A few words about fuel consumption. This has been a big issue among those who believe Earhart had enough fuel to fly to Nikumaroro Island 350 miles to the south.
Other researchers believe Earhart’s fuel analyzer system broke down, and she used more fuel than planned.The Electra’s fuel analyzer system may, or may not, have been working properly. Flying the plane on this route was not all that different from previous flights – at least as far as the engine performance was concerned. Earhart reported no mechanical issues before takeoff and none while flying.
Ric Gillespie, writes in his book Finding Amelia:
“If Earhart used the higher airspeed profile specified in the study (Lockheed’s 487 report), she could fly against an average headwind of 16 miles per hour for the entire flight, arrive in the vicinity of Howland nineteen hours after takeoff, and still have at least a five-hour—26 percent—reserve.”
TEP’s simulator average fuel burn rates were between 55 and 21 gal/hr. depending on what segment of the course we were on. Flying our maximum economy cruise, whenever possible (and looking at our entire flight), TEP ends up with approximately an hour’s worth of fuel at the 19hr. & 12 min. mark. We ended up, consistently, with very little fuel reserve!
We should not forget, too, early in the flight (while still heavy with fuel) Earhart climbed to at least 10,000 feet, and stayed there for some time. That maneuver, we estimate from flying our simulated digital Electra, would have increased their fuel burn rate nearly 24% over cruise flight. We do not know, however, the exact duration of their 10,000-foot excursion.
It should also be noted that as the Electra went into general aviation service it became notorious for its bad cockpit visibility.
TEP discovered it would be hard to spot a mile long island in a sea from 1000 feet at sunrise. TEP has tried it from our model, and the pilot really does not see Howland Island until close in.
According to the USCG the weather over the Itasca was, at the very least, partly cloudy. In other words, the sea’s surface, around the landing site, was not covered by cloud shadows. Howland was in the clear! So why did they have trouble finding Howland? The most likely reason goes back to the USCG weather reports.There were heavy low clouds reported to the north, and west of Howland Island.
Navigation experts, more knowledgeable than those of us at TEP, suggest Noonan may not have been able to get an accurate forward LOP. This may be part of the story, but perhaps not all. Heavy clouds may have forced Earhart and Noonan to descend to 1000 feet as they looked for Howland. Searching at that low altitude for such a small target would not be the best of strategies – especially while traveling up and down a straight line!
The discrepancy between partly cloudy skies over Howland, vs. dark and low clouds to the north and west, may have also confused USCG observers on the Itasca at the time.Those at Howland, under clearer skies, may not have understood the actual conditions our two fliers found themselves in some distance away as they listened to Earhart’s radio calls, and wondered what she & Noonan were doing?
This may be as strong clue as we are ever to get about where Earhart & Noonan went down -- especially if it is combined with our radio geometry observations.
Finally, sitting in the cockpit for 19 hours flying over water, and much of that time in the dark, can produce real fatigue. Earhart & Noonan had been awake more than 24 hours when they believed they had reached Howland Island. Fatigue may have also played a role in their miscalculations. Even in a simulator, 19 hours is a long time!
In any event, something slowly went wrong during the last 200 miles of their flight. Something the fliers did not recognize as fatal – they were lost.
Reviewing the TEP observations after flying the digital Electra L10E “Special” in the X-Plane flight simulator program:
The lack of waypoints during nighttime portion of the flight should not be underestimated.
Potential lack of celestial navigation observations due to possible cloud conditions could produce
uncertain dead reckoning navigation.
The geometry of Earhart radio message signal strengths received by the coast guard may indicate the transmitter was either west or east of the receiver, and probably not directly due north or due south.
Fuel burn rates, and fuel load, most likely do not permit a 350nm excursion to the south after reaching the vicinity of Howland Island.
The digital Electra model, at the end of 19 hours, flew past Howland Island (to the east) on two out of seven flights. An eastward crash site is highly unlikely, but should be considered, and not entirely ruled out.
The Coast Guard reported the skies above Howland Island were partly cloudy. Howland should have been a good target. To the north and west, however, there were heavy, low, clouds that may help explain why they were unable to locate Howland Island from an altitude of 1000’.
The Electra cockpit design guaranteed poor visibility. Crew fatigue may have also been a
serious condition as they tried to find Howland Island. The last 200 miles saw navigation conditions deteriorate, and Noonan’s navigation fail.
BACK TO TOP