Foresight and cool-headed teamwork by pilots of a Royal New Zealand Air Force Boeing 757-200 likely saved their aircraft, 11 fellow crewmembers and 117 passengers—including scientists from New Zealand and the U.S. and New Zealand’s minister of foreign affairs—from a crash-landing in near-zero-visibility conditions at a remote science outpost at the bottom of the Earth in October 2013.

The account is detailed in a New Zealand Transport Accident Investigation Commission (TAIC) “landing below published minima” report, which is accompanied by recommendations aimed at preventing similar occurrences in the future.

The Royal New Zealand Air Force (RNZAF) “40 Sqdn.” operates the 757 and Lockheed C-130 Hercules out of its base in Auckland, providing regular flights between Christchurch and the McMurdo research station in Antarctica as part of a joint logistics pool with the U.S. The arrangement started in 1965, and the 757 joined the fleet in 2010. Whereas the C-130 has several landing options at McMurdo, some requiring skis, the 757 at that time of the season was limited to operations at Pegasus Field and its single 10,000 X 150 ft. runway on permanent ice. The incident flight was the second of three 757 flights scheduled to provide support for the early summer season at McMurdo, with three further flights planned at the season’s end in February and March.

Due to the length of the flight—about 4 hr., 45 min.—procedures called for computing two “point-of-safe-return” thresholds, a worst-case scenario where the aircraft with both engines operating must fly low due to a pressurization problem (PSR-2d), and a less conservative case using both engines at altitude (PSR-altitude). The crew, a captain/instructor, captain-in-training, co-pilot and second captain, decided to use the more conservative PSR-2d, which would occur 2 hr., 47 min. into the planned 4-hr., 40-min. flight. Weather would be a key element in that decision, with the crew receiving real-time and forecast weather over satellite links provided by the U.S. Navy’s Space and Naval Warfare Systems Command’s Office of Polar Programs (SOPP) based in Charleston, South Carolina, and at McMurdo.

The flight had already been delayed two days by weather, and a 7:00 a.m. briefing on Oct. 7 called for “presence of mist with some cloud on the ground,” but with the promise of clearing “during the morning.”

After more weather briefings, the aircraft departed Christchurch at 9:57 a.m. Nearing the PSR-2d time, the crew was becoming increasingly concerned about fog in the vicinity of Pegasus based on updated weather reports. However, the SOPP assured the pilots the fog was “in the distance” and later that it was on the field but was “shallow” and moving away. They were told to expect a visual arrival. The aircraft passed the PSR-2d and, 41 min. later, the PSR-altitude, with forecasters noting that fog was several miles or more away from the airfield.

Weather updates received at 1:45 p.m.—past the point of no return—showed “cloud on ground” with a broken layer of clouds at 300 ft., below the 360-ft. minimum descent altitude (MDA) for the instrument approach to Runway 33. A satellite phone call to the SOPP verified that fog had now formed over the airport.

Committed to a landing attempt, the pilots strategized the arrival, selecting a GPS approach but setting the MDA at 410 ft., due to an unusable section at the beginning of the runway, and asking the airport to turn the approach and runway lights to the brightest setting. The aircraft descended into the cloud bank at 700 ft. altitude and continued down to the MDA, but with no visual sign of the runway, a condition that requires a go-around. The crew climbed to clear air 5,000 ft. and entered a holding pattern for 2 hr. as they waited for conditions to improve, as forecasted. The weather, however, continued to deteriorate, so the pilots set up a second approach, reducing the MDA to 100 ft., an altitude they felt would still provide a safety margin above obstacles around the runway. Passengers and remaining crew were briefed on the situation and seated near the various emergency exits.

At the new MDA of 100 ft. on the second approach, the pilots again did not see the ground; but, during the go-around that followed, a second captain positioned in the observer’s seat in the cockpit saw “several markers and some lights” about 50 ft. to the right of the aircraft. The crew decided that what he had seen were the runway centerline approach lights, meaning the 757 had been about 50 ft. left of centerline on the approach.

The information was somehow familiar—the captain and two other pilots had flown the first flight of the season to Runway 33 in good weather on Oct. 3 and had calibrated the GPS approach and taken pictures. At that time, they had determined that the runway centerline was offset to the right of the inbound course by about 50 ft., a likely consequence of the ice shelf moving since the most recent calibration the previous season. The GPS approach was considered valid, as a new calibration had not yet been conducted.

Using that knowledge and confirming what the second captain had seen against photos that the captain-in-training had taken of the approach on Oct. 3, the crew set up another GPS approach offset to the right by 50 ft. When reaching 110 ft., the co-pilot saw the approach lights and the captain was able to confirm the aircraft was lined up, turn off the autopilot and flare for landing. The 757 stopped with 4,000 ft. of runway remaining and 3,000 kg of fuel, enough for one more approach. The fog cleared 90 min. later.

The TAIC says that, given the weather information and reassurances from the SOPP, the crew was “justified” in continuing beyond the two PSRs and was then left with “no other option” than to land in Antarctica “regardless” of the weather. “The fundamental safety issue arising from this incident was not the decision by the crew to pass the PSR and descend the aeroplane below the MDA in order to make a successful landing,” the TAIC states. “The issue and question is why the aeroplane ended up in that position when the crew had followed standard operating procedures.”

In response, the RNZAF changed its procedures, lowering its cloud base minimums for the 757 flights and adding a requirement that reported or forecast visible moisture below the minimums must be considered, with crews “ultimately being satisfied that the drivers that may cause these conditions to worsen will not be present for arrival.” The 757 procedures were also modified to require that GPS approaches be validated on the first missions of the season.