EDITORIAL

The hunt for MH370

See allHide authors and affiliations

Science  30 May 2014:
Vol. 344, Issue 6187, pp. 947
DOI: 10.1126/science.1255963
PHOTO: STACEY PENTLAND PHOTOGRAPHY

In a world that is increasingly connected, that grows smaller every day, and where so many human actions are exposed to prying eyes, it seems almost incomprehensible that the world's largest twinjet aircraft, with 239 passengers and crew, could vanish for more than 2 months. Determining the crash site of Malaysia Airlines Flight 370 (MH370) has become a scientific detective story, emerging through a combination of scientific technologies used to address problems for which they were never designed. The search for MH370 illustrates a humanitarian dividend from past investments in science as searchers attempt to bring closure to the families and friends of the victims of the tragedy.

A New Zealand military plane searches for debris.

PHOTO: AFP/GETTY IMAGES

MH370 went incommunicado on 8 March 2014. A single Inmarsat satellite exchanged six brief messages with the MH370 Aircraft Communications Addressing and Reporting System. Oceanographers and other mariners have relied on Inmarsat's system of geostationary telecommunications satellites in remote parts of the world's oceans for data and voice communication. Significantly, it is not a navigation system. From the Doppler effect, engineers calculated the plane's velocity relative to the satellite at each time interval, and the delay in the return signal gave them the distance of the aircraft from the satellite. The Doppler was key in choosing the southern over the northern route. Using additional information on the plane's range, they then triangulated to estimate the likely crash site to within 160 km (100 miles). This was the first-ever use of Inmarsat in this mode.

The search area is a remote part of the Indian Ocean. Planes searched for wreckage with no success; ships listened for pings as time ran out on the 30-day batteries sustaining the black boxes. Some promising signals were detected, but the area still to be searched is largely unexplored. The survey tool of choice was a Bluefin-21 autonomous underwater vehicle (AUV). Bluefin Robotics was spun out of the Massachusetts Institute of Technology's Sea Grant Lab. The Bluefin-21 vehicle inherited its distinct construction, gimbal-ducted propeller, mission-control software, and side-scan sonar payload from roots in academic research. Its deep-sea rating that enabled the MH370 response was initially driven by academic applications. The best available bathymetry to help the AUV avoid crashing into rough terrain as it scanned for the debris field of MH370 was assembled from a combination of very sparse ship sonar and satellite altimetry. Satellite altimeters, first launched nearly 40 years earlier to map the ocean surface, produced better maps of the seafloor than were available from shipboard echo-soundings alone, by using small-scale features in the marine geoid to estimate the shape of the seafloor. Scientists had been using these maps to better understand Earth beneath the ocean; now the map would help guide the AUV in a search area still needle-in-a-haystack large and more than 4000 m deep.

As of this writing, the search continues. In April 2011, a team from the Woods Hole Oceanographic Institution found the debris field from Air France Flight 447, which had crashed into the Atlantic Ocean 2 years before, after a week of searching with similar AUV technology,* bringing resolution to the families of the victims. Finding the black boxes is vital to avoiding similar incidents happening again. We hope for the same outcome for the MH370 search. But it took 2 years to narrow the search to the right part of the Atlantic. In both cases, the response could have been improved by filling known gaps in scientific understanding (see the News story on p. 963). For example, the resolution of the satellite-derived map guiding the Bluefin-21 is ±250 m vertical and 15 km horizontal. Relative to the plane's dimensions, the unknowns are serious. For comparison, the resolution of features on Mars is ±1 m vertical and 1 km horizontal. And knowing where to look saves precious time, whether one seeks a plane full of passengers or a truck full of Nigerian schoolgirls. We should use better technology to track what is too valuable to lose.

Navigate This Article