We turned right and lined up on the inbound course for the ILS approach, still able to see the frozen ground out the side windows. But looking straight ahead, we saw nothing except white snowy gloom. The end of a near-maximum-range leg over Greenland beckoned, and we hoped for a warm FBO lounge, a quick load of fuel, and no need to spend an unplanned night in far northern Nunavut.

The TBM 910’s dependable 850-shp Pratt & Whitney Canada PT6-66D was running perfectly, having propelled us from our departure the previous day from Daher’s Tarbes, France factory to our next fuel stop at Iqaluit, the capital of Nunavut on Canada’s Baffin Island, perhaps more familiar to some as Frobisher.

We were roughly halfway to the destination on this ferry flight to deliver a new TBM 910 to Daher distributor Avex in Camarillo, California, and already I had developed a keen appreciation of the challenges of ferry flying. Gilles Glatz, who flies as many as 30 to 40 of these kinds of trips per year, had kindly assented to my joining him on this one, perhaps the most challenging of its type, because it was the middle of winter in the Northern Hemisphere. Daher’s Philippe de Segovia wanted me to experience the true performance capabilities of the TBM 900 series—I had flown one a few years ago for a pilot report—and what better way than a delivery flight from the Tarbes factory to California?

The logistics of such a trip are daunting, but Glatz has done this many times, and Daher is well experienced at preparing TBMs that roll squawk-free off the assembly line, ready for long-distance delivery flights. De Segovia said that the trip should take three to four days, and it was arranged that I would depart the U.S. East Coast for France on a Sunday and arrive on Monday afternoon. Because of the outlandish prices for flights to Tarbes, I chose to fly to Toulouse and rent a one-way car to drive to Tarbes.

The travel was slightly complicated by my having to pack an extra bag of winter clothing, which Glatz had recommended (and that I was glad I'd brought). It took a large extra suitcase to fit a pair of hiking boots, snow boots, my Carhartt winter jacket (the brand that scientists and technicians wear during winters in Antarctica), three pairs of gloves, long underwear, two winter hats, a down jacket, and a thick sweater.

I arrived on a cloudy and drizzly day and found my way onto the highway to Tarbes. Not having slept much on the overnight flight from New York, I struggled to stay awake and stopped at a roadside cafe for coffee, once again stunned at the high quality of food in such places, compared with the crappy food we Americans put up with.

Arriving at Tarbes, De Segovia offered a bag of baguette sandwiches that he had procured for us, but as I had arrived later than expected, I put off eating, and this later proved highly beneficial. He gave me a tour of the vast Daher factory complex, where the company builds not only the TBMs but also major structural components for Airbus (airplanes and helicopters) and Dassault (I’m pretty sure I saw some Falcon 6X components there, in advance of Dassault making the formal announcement of the 6X program on February 28).

Then it was time to climb on board N910GE, a new 910 model with only 4.7 hours’ total flight time. The exterior is painted a handsome two-tone gray with a sporty yellow stripe, and the interior features matching yellow stitching on the optional gray leather seats. The interior was still protected with floor coverings and plastic over the seats. Although the turboprop carries six people, our luggage took up all of the 30-cubic-foot baggage compartment and some space in the cabin. Glatz brought a life raft and two survival suits along with his own winter wear. In addition to my luggage, I had a spare external Bad Elf GPS receiver, a Spot satellite emergency beacon, a handheld marine radio, and binoculars. The latter two items were for help finding ships in an emergency, as I was trying out a new iPad app called Ergo 360, which shows locations of ships on over-ocean legs. For tracking our flight, I used the ForeFlight app, which has an excellent world basemap. Glatz uses Jeppesen Mobile FliteDeck for charts.

Of course, all the information we needed was available on the TBM 910’s Garmin G1000 NXi avionics suite, which features the latest G1000 displays with faster processors and HSI map with multiple overlay options such as weather, terrain, and traffic.

On the instrument panel, the two 10-inch GDU 1050A primary flight displays flank a 15-inch GDU 1550 multifunction display. This TBM 910 had a long list of standard equipment and many options installed, including synthetic vision, GSR 56 Iridium datalink, GWX 70 digital radar, RVSM (although not available to us because the new owner would apply for the RVSM letter of authorization), ADS-B Out/In, Mid-Continent MD302 Standby Attitude Module, and much more.

The TBM 910 is filled with creature comforts: cupholders in the cockpit, USB ports on the panel in front of the right pilot seat, a single-lever power control, automatic fuel tank switching, three-axis electric trim, electrically adjustable rudder pedals, and automatic pressurization and environmental control. 

With a maximum-fuel payload of 891 pounds, we had plenty of capacity when taking off with a full load of fuel (291 gallons or 1,980 pounds). Published range for the TBM 910 is 1,730 nautical miles, but that is at a slower speed, 252 ktas, while carrying one pilot and landing with a 45-minute reserve. At maximum cruise speed (326 ktas) at FL280, a single pilot should be able to fly 1,200 nautical miles with NBAA reserves (100-nautical-mile alternate). If able to fly higher at the maximum altitude of FL310, that range would climb to 1,350 nautical miles.

Day 1: Tarbes to Prestwick

Glatz is not one to waste time, and at 5:02 p.m. we taxied past rows of aging airliners stored at the airport. Then he pushed the power lever forward and took off in cloudy but clearing weather.

The 783-nautical-mile flight to Prestwick, Scotland, took us over Cherbourg, France, at FL280 in the evening light that was still brightening the cloud layers accompanying us as we flew north across the English Channel toward Portsmouth.

With fuel nearly full, our takeoff weight was just under 7,000 pounds, about 400 pounds below the 7,394-pound mtow, and we climbed at 150 kias, burning 70 gph in the climb. We leveled briefly at 14,000 feet, then resumed climbing and picked up a faint trace of ice on the booted leading edges passing through 17,000 feet, where the outside temperature was -20 degrees C. Once we reached FL280 (OAT -41 degrees C), the airplane’s optimal cruising altitude for the highest we could go without RVSM approval, cruise speed initially settled at 301 ktas, and fuel flow dropped to 55 gph. Cabin altitude at FL280 was 8,400 feet.

The TBM can fly faster at this altitude, more than 320 ktas, Glatz explained, but it’s not worth burning the extra fuel. He generally picks the altitude (FL280 when possible) and sets the fuel flow, depending on whether the wind is on the nose or tail. If a strong tailwind is blowing, he will pull back the fuel flow to 45 gph. Our RocketRoute flight plan estimated 3:14 flight time and 199 gallons of fuel burned.

The headwinds initially blew 80 knots on the TBM’s nose, but over the Channel these dropped to the low 30s. The setting sun lit up the western horizon with a rainbow of orange hues deepening to purple as we descended into Prestwick, where the wind was calm and cloud layers ranged from 10,000 feet scattered to 15,000 feet broken. Glatz lined us up on the ILS to Runway 30, and we landed at 3:08 after departing Tarbes, having burned just 148 gallons.

The delightful and inexpensive Carlton Hotel awaited us once we cleared customs—thankfully the customs officer came to the FBO—and Glatz happily pointed out that we were much better off having flown this first leg on Monday evening instead of spending a potentially sleepless night in Tarbes, then trying to get an early start. Otherwise, we might have risked a possible overnight in a tiny, frozen outpost in Greenland.

Day 2: Prestwick to Winnipeg

A full Scottish breakfast was an excellent way to start what turned out to be a long day of flying. Ferry pilots must be ready to take advantage of good weather.

Glatz had arranged the previous night for a taxi to meet us at 7:30 a.m., and I could tell he was getting a little anxious when it didn’t arrive exactly on time. The driver did show up a few minutes later, and after a short ride to the airport, the friendly staff at the Executive FBO helped us load the TBM. They had already filled it with fuel, so once again, it was a prompt launch with no delays, this time for what should have been a relatively short hop to Reykjavik, Iceland.

The weather was benign, considering that we were flying across the Atlantic Ocean at the worst time of the year. What would challenge us on the first leg of the day were extremely strong headwinds. Our routing from Prestwick to Reykjavik was 751 nautical miles, but the wind would be right on our nose, and unavoidable if we wanted to make headway on this journey.

This impressed on me yet again the unrelenting constraints that come with ferry flying. In this part of the world, we had some airports to choose from if we had to stop for fuel before the ocean crossing; the Faroe Islands was one such alternative, but that is a location with what Glatz characterized as skimpy services, and he avoids landing there unless absolutely necessary.

As it turned out, even with massively strong headwinds, the TBM had the range to make Reykjavik with decent reserves, and it’s a good thing, because as is typical with headwinds, they were worse than forecast.

We departed Prestwick as the sun began warming the wet ground, not wasting any time getting moving once loaded up.

On this leg, traffic was more sparse, and we were quickly cleared to FL280. Once past Stornoway and feet fully wet, we were cleared direct to barku intersection, and now the wind did its thing, howling from the northwest, churning up the sea far below and relentlessly forcing our groundspeed numbers down into the low 200s. Glatz pulled the power back, lowering fuel consumption to 50 gph, but groundspeed dropped to 219 knots, so he pushed the power up to 56 gph, and while the true airspeed crept up, groundspeed remained the same. About 50 miles shy of barku, the wind reached its peak at 126 knots from 297 degrees, which translated to 116 knots on the TBM’s nose. The wild ocean below seemed to be moving by us extremely slowly.

Watching the fuel over destination (FOD) number on the G1000 moving map helped illustrate the effects of the wind and our attempts to mitigate those effects as much as possible by descending, climbing, and descending again. Nothing really helped, and we ended up settling on FL260, where the FOD was 89 gallons, about 15 gallons better than at FL280. Going any lower would have increased fuel consumption without adding speed, so we were in the most efficient space possible.

The TBM soldiered on while I tested the ADS Ergo 360 ship-mapping program on my iPad. I had downloaded a fresh load of ships at the FBO just before takeoff, and now I practiced pulling up the ship information, then plugging their lat/long positions into ForeFlight and creating a user waypoint that we could navigate to. There were a surprising number of ships, mostly cargo and fishing vessels, plying their trade in the chilly North Sea.

During this leg we were in the clear with cloud layers below so I couldn’t spot any of the ships that I was monitoring, but knowing they were there was a big comfort. Although the windspeed dropped slightly after we turned left at barku, now it was even more aligned with our flightpath. Groundspeeds remained frustratingly low.

Flying long distance involves a lot of straight-and-level with the autopilot on, monitoring the airplane and, in this case, managing the all-important fuel resource. Glatz doesn’t like to eat much while flying; he does so much ferry flying that he feels he would just be snacking constantly if he allowed himself to indulge. I, however, had no qualms about digging into the food bag that de Segovia had so kindly provided, and it contained three wonderfully tasty chicken and cheese sandwiches, one of which had been yesterday’s dinner, and the second today’s lunch. Glatz had brought a thermos of coffee and shared it with me as the wind tried to hold us back, eventually slowing us to 194 knots groundspeed. Even so, Glatz retained his optimistic viewpoint, noting that if we were in anything other than a TBM, we might not have been able to make the Prestwick-Reykjavik hop in those conditions.

Finally, after three hours and 47 minutes, we skimmed over the snow-dusted rooftops of Reykjavik and touched down on an ice-free Runway 19 in excellent visibility and high clouds, still with nearly 80 gallons of fuel on board. Not that there are many alternates in this part of the world, but this is the reality of ferry flying. Without wind, the flight would have taken 2:39.

We spent less than 40 minutes fueling up, checking weather, and clearing customs in Reykjavik. Glatz suggested I put on my long underwear to prepare for landing in Canada. Sadly, we had no time to spend a night at the airport’s decadently comfortable former Loftleider Hotel (now the Icelandair Hotel Reykjavik Natura) with its fluffy pillows and duvets and geothermal heat, but Glatz said prices had climbed to outrageous levels with the recent growth in Iceland tourism, and in any case, we needed to make tracks while the going was good, which it was for the most part.

For the next leg, the wind wasn’t such an issue, and it appeared that over Greenland we would even pick up some tailwinds, according to the handy Windy.com app and website that Glatz uses. This would help enormously, as we were about to fly our longest leg, 1,207 nautical miles over Greenland to Iqaluit on Baffin Island, where the forecast was not bad except for gusty winds and blowing snow cutting visibility on the surface.

I did the takeoff from Reykjavik. Glatz suggested bringing the torque to near 90 percent while rolling, then keeping it below 100 percent with small adjustments. The TBM accelerated quickly and I soon pulled back on the yoke and as the speed built up, retracted the landing gear, then flaps. The TBM has comfortable handling, but trim is needed to keep pitch forces from getting too heavy. Turning to the west, I trimmed, climbed at 150 kias, then switched on the autopilot as we neared 10,000 feet.

It wasn’t long before we crossed the east coast of Greenland, north of the airport at Kulusuk, although there wasn’t much to see because of clouds below. We started out with headwinds, much lower now that we were flying more westerly, and 100 nautical miles from Greenland the wind shifted to our tail, giving us a much-needed push toward our destination.

As we flew farther west, the groundspeed gradually increased as the wind dropped, and the FOD number kept climbing into the eighties. We were able to pull the power back and lower fuel consumption to 45 gph, and even at 274 ktas our groundspeed climbed to 307 knots.

The groundspeed stayed comfortably high as the mountainous icy surface of Greenland passed beneath, just 17,000 feet below us, according to the agl readout on ForeFlight.

The OAT was quite cold this far north, in the minus 50s (centigrade), and this might have been why we suddenly heard a keening noise from somewhere in the flight deck. It sounded like an air leak. A look at all the engine and pressurization gauges showed all systems nominal, but the noise persisted.

Glatz decided we should descend and try to find warmer air, and perhaps that would help with what we figured was a leaking seal on the TBM’s pilot door. We asked ATC for FL260, and I dialed the autopilot down for the descent. The temperature situation improved, but it didn’t solve the leak, which was annoying, even with our Bose A20 active-noise-reduction headsets. Finally Glatz did a quick on/off cycle of the pressurization system, and that took care of the leak and the sound stopped. For flying in such cold temperatures, the TBM pilot information manual recommends applying silicon grease to door seals, probably to prevent just such an occurrence.

The lower layers of cloud began to break into towers, leaving clear views of the Greenland landscape, which was utterly devoid of any human activity. We flew over giant steep mountain passes buried under mounds of ice and snow, and flat fields that looked snowy soft from up high but that likely hid deep crevasses and unyielding boulders. The terrain dropped lower and lower as we flew over the west coast, and Glatz pointed out one of the few airports just a dozen miles north of us; it was Sisimuit, and we could see the 2,600-foot runway, free of ice and snow and black against the surrounding endless whiteness; a good emergency field, but not suitable for regular stops, as ForeFlight’s information indicates the airport has no fuel or other services. So far the wind and groundspeed still made Iqaluit a good stop.

Glatz took over about an hour from Iqaluit as conditions there were going to require an approach to near minimums. We kept flying high as long as possible to maximize endurance; the TBM is a champion at descending quickly when necessary.

One benefit of flying so far off the beaten path is little or no traffic. ATC cleared us straight to the final approach fix for the ILS Runway 34, and a simple 90-degree turn put the TBM on final. The visibility wasn’t bad, and we could see terrain out the side windows, but looking ahead was somewhat murky. The approach lights materialized from the snowy gloom just before we reached minimums, pointing the way toward the welcome runway. The blowing snow made the runway hard to see, but there was plenty of visibility while looking ahead along the runway. Glatz made what must be one of the shortest TBM landings in the 24-gusting-to-30-knot wind, which was blowing almost straight down the runway. The 4:25 flight was our longest leg, but we still had 70 gallons of fuel, plenty to make it to the few airports on Baffin Island.

We taxied to the FBO ramp in the blowing snow, and while Glatz was shutting down the engine and calling the fuel truck, I climbed in back and prepared to go outside in the -27 degree C gusty weather. I put on a sweater, then my down jacket and my Carhartt winter jacket on top. On my feet I wore two layers of socks—one a thick winter pair—then I stuffed my well-insulated feet into my winter boots. My head was covered with a wool balaclava and the down jacket hood, and I wore thick gloves.

When we opened the door—after Glatz put on his winter outerwear—the wind sucked the heat right out and almost blew me over when I climbed onto the snowy ramp. The FBO office beckoned, but oddly we could see no lights. We trudged through thick snowdrifts to the door, and it was locked. No warm office for us.

The fuel truck pulled up, and two courageous linemen dressed in thick full-body cold-weather suits quickly topped off the TBM. We had given up on finding anything open on the airport, except for the flight service station, which we deemed too far to walk to, so we climbed back onboard after Glatz double-checked the fuel caps and paid for the fuel.

We had been told in Iceland that we would need to call Canada’s border service agency—via the CanPass service—once we arrived, to gain clearance, as there were no officers at Iqaluit that day. Unfortunately, none of our telephones—Glatz had two, each with different service providers—was able to capture any cellular signals. This was the first place in the world where my T-Mobile-powered iPhone wouldn’t connect.

By great good luck and a bit of forethought, Glatz had logged in to the FBO’s Wi-Fi network on his last trip through Iqaluit a few months earlier, and more luck, the password had not been changed, so he was able to connect and make a Wi-Fi call on his T-Mobile phone to CanPass and get us cleared into Canada.

After starting the engine, Glatz taxied back to the runway, and we waited for a break in the blowing snow before advancing the torque and pushing on into the heart of frozen Canada. We did not want to dally in Iqaluit, he pointed out, because if we stayed there too long, we’d run the risk of the airplane icing up and cold-soaking the engine, in which case we’d have to stay overnight and pay to hangar the TBM. At Iqaluit, he said, it cost several hundred dollars to open the hangar door and the same amount to close it. An overnight stay would cost a minimum of two openings and closings. Not that there was any pressure to pit safety against cost; Glatz is paid regardless, and the manufacturer normally covers such operational expenses. But our mission was to deliver the TBM to California, and unnecessary overnight stays were not productive.

We had hoped to make it 1,241 nautical miles from Iqaluit directly to Winnipeg, where Glatz’s friend Dylan Fast owns the Fast Air Jet Centre FBO. But flight planning scenarios showed that we would arrive with little extra fuel if the wind wasn’t favorable. A direct leg to Winnipeg would also mean an extra-long flight across the frozen Hudson Bay with no nearby airports. A better option would be to fly more westerly to Churchill, just 792 nautical miles, then the 542 nautical miles to Winnipeg. If the winds proved helpful, it could have been possible to turn the corner at Churchill and make it to Winnipeg; otherwise a stop at Churchill looked likely.

With the airplane leveling at FL260, the OAT was still cold, and the wind was back on the nose; it was obvious early on that we would land at Churchill. Hudson Bay is massive, and it took what seemed like endless hours to cross the frozen lumpy expanse. Although it was midwinter, long open cracks in the ice zig-zagged tens of miles into the distance. The visibility was nearly unlimited on this leg, but there was little to look at outside other than the sun shining on the white iciness far below.

The winds started in the 20s, but not on the nose so it didn’t have a huge effect. We had dropped lower to FL240 because of the low OAT, and pulled the power back to the 45-gph setting, for a true airspeed of 257 ktas. As we crossed the western side of Hudson Bay, the wind shifted back to the west and slowed us down. Back at FL260, with a true airspeed of 283 knots, groundspeed was a low 227 knots.

It was late afternoon, and the sun shone brightly on our little airplane as we traversed the vast emptiness. Finally it was time to descend into Churchill, and we skimmed over a light fluffy layer of stable wispy clouds, which suddenly enveloped us in a clammy embrace, forcing us to remove our sunglasses and adapt to the gray-lit landscape below.

The flight to Churchill took 3:28, and thankfully the FBO was open. It turns out we were lucky to get fuel, as the railroad tracks that serve the airport had been damaged in a storm, and trainloads of fuel were unable to get through. In any case, we landed with more than 100 gallons and would have had plenty to make it to one of the other airports southwest of Hudson Bay.

One more short leg, and we would call it a day. We spent as little time as possible at Churchill, where the OAT was a balmy -21 degrees C, much more comfortable and less windy than Iqaluit.

It was just another 2:18 to Winnipeg. After a quick climb to FL260, the airspeed settled at 292 ktas and groundspeed 248 knots, burning 55 gph. The door seal did its cold-OAT thing again, but Glatz took quick care of it, and that was the last we heard from the seal on this trip.

After what seemed like hours of emptiness, occasional signs of civilization appeared. Tiny towns, many with airports, gradually passed by as we motored on into the sunset. At 100 nautical miles from Winnipeg, ATC asked us to descend slowly at high speed, so we tipped the nose into a 1,000-fpm descent, and the leading edges of the wings lit up as if we were flying so fast that the air molecules were sparking into tiny orange burning embers.

It was dark by the time Winnipeg appeared, and the sight of the bright runway lights was most welcome after a challenging day. We had logged nearly 14 hours of flying and made it from Prestwick to Winnipeg in one day. The next day's trip should be relatively easy, and the remaining challenge was clearing customs in the U.S.

Glatz and I celebrated the long day’s work with a well-deserved steak-and-beer dinner at The Keg Steakhouse, and Fast showed up later and entertained us with Canada flying stories.

Day 3: Winnipeg to Camarillo

While the last day’s distance was much less challenging, Glatz did want to try to get to Camarillo in time to catch a taxi to Los Angeles International Airport for a flight to Geneva, where his next assignment on his endless journey around the planet awaited. The TBM was fueled and ready, so we took off soon after arriving at the airport and headed for the border and our customs stop at Williston, North Dakota.

This was a relatively short leg, just 273 nautical miles, and took less than an hour and a half. Glatz is an expert at importing aircraft into the U.S., and he had made all the arrangements ahead of time, so the customs process at Williston went smoothly. We didn’t need fuel and took off planning to stop somewhere in Utah, depending on the winds and fuel prices, which I could handily view on the ForeFlight app.

With so many airports to choose from, it was almost hard to make a decision, but we ended up able to make it to Fillmore, a small town 100 nautical miles south of Salt Lake City. I took the controls for the landing on Fillmore’s 5,000-foot Runway 04. Although I came in too high, the TBM loses altitude easily, and I pulled the power and lowered the nose until we were on the proper glidepath, then added a little power and flew down to the runway and a decently smooth touchdown. The TBM’s handling is precise but somewhat heavy, although lighter at slow speeds. As with most high-performance airplanes, trim helps the pilot manage pitch control forces.

Fillmore is one of those typical barely used airports that dots the U.S. landscape, populated with a few airplanes and a hangar trustingly left unlocked for visiting pilots. The fuel is all self-serve. We topped off the tanks, adding a can of Prist to each one, enjoying for the first time warm sunny weather and the unobstructed view of the snow-capped mountains to the east.

I did the takeoff from Fillmore, and it was a pleasant way to start the final leg of the trip. We had plenty of fuel and calm winds, so we didn’t need to climb all the way to FL280 and settled at FL240 with 35-knot headwinds. Glatz took advantage of the opportunity to climb back into the cabin and pack up his bags while I dealt with ATC and a clearance change that was easy to input in the G1000 NXi flight plan.

All too soon, it was time to descend, and Glatz took the controls for the remainder of the leg. Moderate turbulence bounced us around as we flew over Barstow and Victorville, through the Mojave Desert and over Palmdale, past Edwards Air Force Base through Simi Valley—familiar territory from when I lived in Los Angeles.

The turbulence finally calmed as we descended toward Camarillo Airport, and Glatz brought the TBM to a gentle landing on Runway 26, just 23.5 flying hours from our departure in France only two days previously. It was hard to believe we had flown 5,400 nautical miles in such a short time, but that is the life of a ferry pilot.