Researchers have found Amelia Earhart's last picture, taken the day before her fateful flight over the Pacific.
Happy and smiling, Earhart is shown in Lae, New Guinea, as she waits for her aircraft to be fueled following a test flight on the morning of July 1, 1937.
In the picture, she is standing with Frank Howard, the local representative for Vacuum Oil who was supervising the fueling.
The following morning, Earhart and her navigator Fred Noonan took off for the longest and most difficult leg of Amelia's record attempt to fly around the world at the equator.
RELATED: Amelia Earhart's Fate Reconstructed: Photos
"It was meant to be a 2,500-mile jump to Howland Island, a tiny coral outcropping in mid-Pacific where a runway had been carved out," said Ric Gillespie, executive director of The International Group for Historic Aircraft Recovery (TIGHAR), which has been investigating Earhart's fate for nearly three decades.
The Coast Guard cutter Itasca was waiting there to refuel Earhart's twin-engine Lockheed Electra for the flight to Hawaii.
"From there she would fly to Oakland to complete the world flight," Gillespie said.
But Earhart and Noonan never arrived at Howland Island.
Earhart's last messages were heard by the Itasca on the morning of July 2nd, as she flew toward her target destination. One message, send at 07:42 local time, explained that she was low on fuel.
"We must be on you, but cannot see you -- but gas is running low. Have been unable to reach you by radio. We are flying at 1,000 feet," Earhart said.
RELATED: Amelia Earhart's Plane Discovered in Odd Movie Cameo
Her final inflight radio message occurred a hour later, at 08:43.
"We are on the line 157 337. We will repeat this message. We will repeat this on 6210 kilocycles. Wait," she said.
According to TIGHAR, the numbers 157 and 337 refer to compass headings -- 157 degrees and 337 degrees -- and describe a navigation line that passed not only Howland Island, the target destination, but also Gardner Island, now called Nikumaroro.
This uninhabited atoll in the southwestern Pacific republic of Kiribati is where TIGHAR believes Earhart and Noonan landed safely and ultimately died as castaways.
"An abundance of archival, photographic and artifact evidence suggests that Earhart and Noonan made a successful landing on the island's fringing reef," Gillespie said.
RELATED: Earhart's Anti-Freckle Cream Jar Possibly Found
In a video presentation on YouTube last month, Gillespie explained that Earhart sent radio distress calls for nearly a week before the aircraft was washed into the ocean by rising tides and surf.
Using digitized information management systems, antenna modeling software, and radio wave propagation analysis programs, TIGHAR re-examined all the 120 known reports of radio signals suspected or alleged to have been sent from the Earhart aircraft after local noon on July 2, 1937 through July 18, 1937, when the official search ended.
They concluded that 57 out of the 120 reported signals are credible.
"The post-loss radio signals are actually the one body of evidence that cannot be explained away by even our most dedicated critics," Gillespie told Discovery News.
"There is a growing realization that there is no reasonable alternative interpretation for the 57 credible post-loss radio signals that were heard in the days following Earhart's disappearance," he added.
RELATED: Amelia Earhart Letter Encourages Aspiring Aviator
To make multiple transmissions, the Electra plane needed to be on land and on its wheels. It would have also needed to run the right-hand, generator-equipped engine to recharge the batteries.
"The safest procedure is to transmit only when the engine is running, and battery power is required to start the engine," said Gillespie. "To run the engine, the propeller must be clear of obstructions and water level must never reach the transmitter," he added.
TIGHAR analyzed tidal conditions on the island from July 2-9, 1937, the week following Earhart disappearance.
It emerged that transmission of credible signals occurred in periods during which the water level on the reef was low enough to permit engine operation. At least four radio signals are of particular interest, as they were simultaneously heard by more than one station.
The first signal, made when the pilot had been officially missing for just five hours, was received by the Itasca, and two other ships, the HMS Achilles, and the SS New Zealand Star.
The Itasca logged "We hear her on 3105 now -- very weak and unreadable/ fone" and asked Earhart to send Morse code dashes.
The Achilles did not hear "very weak and unreadable" voice, but heard Itasca's request and heard dashes in response. The SS New Zealand only heard the response dashes.
In other cases, credible sources in widely separated locations in the United States, Canada and the central Pacific, reported hearing a woman requesting help. She spoke English, and in some cases said she was Amelia Earhart.
RELATED: Hidden Moon Crater Named After Amelia Earhart
In one case, on July 5, the U.S. Navy Radio at Wailupe, Honolulu heard a garbled Moorse code: "281 north Howland -- call KHAQQ - beyond north -- won't hold with us much longer -- above water -- shut off."
At the same time, an amateur radio operator in Melbourne, Australia, reported having heard a "strange" code which included KHAQQ, Amelia's call sign. According to TIGHAR, such distress calls were ignored.
"Earhart and Noonan eventually died as castaways on the waterless, uninhabited atoll, their aircraft washed into the ocean," Gillespie said.
After 11 expeditions to Nikumaroro, Gillespie is now trying to raise money for a 12th in the summer of 2017, on the 80th anniversary of the disappearance of the legendary pilot.
The search for whatever remains of Earhart's Electra will rely on two, three-person manned submersibles operated by the University of Hawaiii's Undersea Research Laboratory (HURL). They will inspect the underwater area down to a depth of up to 6,500 feet.
Gillespie estimates the expedition will cost $1,750,000.
See Photos: Amelia Earhart's Fate Reconstructed: