Art Courtesy of Kara Tao.
It was March 8, 2014—a day like any other—when 239 people took to the skies aboard Malaysia Airlines Flight 370 on their way from Kuala Lumpur to Beijing. Some were going home after a long time away. Others were world-famous calligraphers returning from a business trip. Some may have been scared of flying and clutched the armrests as the plane took off. But after that fateful day, none of those 239 people, nor the plane they sailed away on, were ever seen again. And despite years of intensive searching—using everything from submarines to sonar imaging—their final resting place has yet to be discovered.
Over a year later, on July 29, 2015, Gregory S. Herbert, Associate Professor of Paleobiology at the University of South Florida, was watching the news and saw that a piece of the missing aircraft’s wing, called a flaperon, had been found on Réunion Island. Herbert instantly knew that he had to make some calls. A clue that could unlock the location of the lost plane had been unearthed, and he was uniquely qualified to decode it.
Herbert’s background lies in stable isotope geochemistry; specifically, he decodes ocean temperatures from barnacle shells. If a drifting object has barnacles, scientists can potentially use these temperatures to track its path through the ocean. And barnacles, clinging to the flaperon, were clearly visible on the TV screen. “I knew immediately that there were sea surface temperatures recorded in those barnacles,” Herbert said. “Some of the barnacles were fairly large, and they could have recorded the whole drift.”
Herbert tried to contact the French authorities, who had possession of the flaperon, and the Malaysian officials, who were running the investigation. Both attempts failed. However, Herbert was not deterred, and the third time proved to be the charm: the Australian authorities, who helped coordinate the search since the plane’s likely final location nears their territory, enthusiastically agreed to look over his proposal.
Based on satellite data, the plane’s final resting place is thought to lie somewhere in the Indian Ocean along the seventh arc, between latitudes twenty and forty degrees S. However, this is an extremely large area that the plane may not even be in. But with the technique Herbert and his colleagues have developed, scientists can say for sure whether the plane is in the seventh arc, and can pinpoint its location to a smaller and more easily searchable area.
Barnacles grow in daily layers, similar to the rings trees produce every year. Each of these layers encodes chemical data about their surroundings at the time of growth. Different isotopes of oxygen are deposited at different sea surface temperatures, with a known relationship between their ratio and the temperature. Scientists can analyze this ratio through δ18O values to determine the temperature the barnacles experienced each day, and match that data with different temperature currents that run through the Indian Ocean. Other scientists had previously jumped on this information to produce temperature and location models for the aircraft, but in their rush to complete the work, they failed to use experimental controls, leading to large uncertainties in their results.
Despite these apparent problems with the previous studies, Herbert had a difficult time securing funding for his study. In the end, the Florida Aquarium decided to fund his research, as it could also be used to benefit sea turtles. Sick sea turtles will float for weeks and thus develop barnacles on their normally clear front flippers. If these barnacles could be traced, scientists could begin to identify areas where sea turtles tend to get sick. Thus, a method was born that could both trace a missing plane and track sick turtles.
This new technique, created by Herbert and his colleagues, had two unique and vital components that set it apart from previous attempts. The project was the first to create an experimentally derived equation for the particular species of barnacle (cosmopolitan stalked barnacle, Lepas anatifera) that was attached to the flaperon. Barnacles were placed into tanks, stained with a marker (a fluorescent dye) that showed divisions between layers, and subjected to slowly changing temperatures. The scientists then anesthetized the barnacles and analyzed their layers for δ18O content. Finally, they created an equation that relates temperature and δ18O content.
The second innovation centered around what to do with that temperature data. While temperature does vary throughout the ocean, there are large bands that are the same temperature throughout. “Just knowing that first temperature doesn’t tell you where the plane is; you have to do a lot more work,” Herbert said. In other words, each new temperature recorded is needed to narrow down the search; knowing just the first temperature recorded by the barnacle is not enough.
This extra work involved developing a modeling simulation using known sea surface temperatures and other data such as current velocity that is consistently recorded across the oceans. The simulation allowed the researchers to cast virtual flaperons adrift from various starting points, and then statistically analyze their routes to determine the most likely path each barnacle on each flaperon took based on its temperature data.
Herbert and others on the team applied this technique using previously published data for one of the smaller barnacles found clinging to the flaperon. First, they calculated the barnacle’s age at each layer through an experimentally derived equation relating barnacle size to age. “We measured the size of the barnacle at each sample, at each temperature,” Herbert said.
They then cast 50,000 virtual flaperons adrift in different places along the band of the ocean defined by the barnacle’s earliest temperature value. Then, they compared the temperature data these virtual flaperons experienced with the actual temperature data from the barnacle using a method called dynamic time warping. Eventually, this eliminated all but one virtual flaperon, which was the only one to end near where it was actually found: in waters near Réunion Island.
However, the timeline for applying this new and promising technique will have to wait on the French government, which has custody of the largest barnacles. These are the only barnacles that could have recorded the entire drift of the flaperon. “I have a feeling that they’re still sitting on these shells because there were three French scientists who worked on them, and their work was very rushed, and they did not get any sort of a conclusive result,” Herbert said. The French government likely wants to keep the samples until the foundational work that will allow for conclusive results has been completed. Thus, it is possible the French government will release the barnacles in light of these new findings.
In the meantime, the next step is to improve the model and equations. “I just wanted to demonstrate how to do the method first,” Herbert said. To begin, Herbert and others have already started work on a more accurate barnacle age model, since shell size is not the most accurate predictor. They also want to improve the flaperon motion model used; for example, accounting for the fact that a flaperon does not behave like an idealized buoy, and instead drifts slightly left. Finally, the researchers need to perform a sensitivity analysis. This involves running the model thousands more times with different errors factored in to see how dramatically these errors change the results. This work would take up to a year, even if the larger barnacles from the French were provided immediately.
However, hopes are high. Of the five drifters in the simulation that best matched the known barnacle’s path, four of them started in the same location, tightly clustered together. “We’re not just looking for a single temperature, we’re looking for a sequence, a very unique sequence of temperatures. And there aren’t that many drift origins, and drift pathways, that could possibly be consistent with that,” Herbert said. When asked if the plane would ever be found, Herbert didn’t hesitate.
“Yes,” he said.