Written by experienced pilot Patrick Smith, this blog addresses commercial air travel issues pulled from the headlines, including crashes, on-board service and innovations. With his personal essays and travel stories, Ask the Pilot offers a unique perspective on airline issues from around the world.
ONCE AGAIN it’s the first day of July — the anniversary of the day, in 1986, when I was nearly killed in a small-plane collision over Nantucket Sound.
In interviews I’m often asked about my most memorable or frightening close call. Apparently a lot of people dig hearing pilots talking about being scared or screwing up. The story below, which I post every year on this date, is my response to that question.
The fact that my answer takes us back more than three decades, to when I was but a 20 year-old private pilot at the controls of a four-seat Piper, should underscore the incredible safety of flying in general, and especially of commercial flying.
With me in the airplane that afternoon was a young girl with whom I long ago lost touch, and whose beauty and peculiarities are detailed in the story that follows. I have no idea what her own recollections of that day might be, but mine remain crystal clear.
THERE HAS BEEN a lot of buzz of late surrounding Chernobyl, the HBO miniseries about the 1986 nuclear disaster in the Soviet Union.
In April of that year, reactor four at the Chernobyl nuclear power plant, in present-day Ukraine, exploded, sending plumes of radiation across Europe in what is still, by far, history’s worst nuclear accident.
To this day, a 30-kilometer “Exclusion Zone” surrounds the site, accessible only to researchers, temporary workers, and a small number of villagers — most of them senior citizens — that the Ukrainian government allows to live there. And, believe it or not, to tourists. Day trips to Chernobyl can be arranged in the capital, Kiev (Kyiv), and include transportation to and from the site, plus all the admission formalities — and a radiation scan on your way out.
In 2007 I took one of those trips. The photographs below are from that day.
I have not captioned the pictures. They more or less speak for themselves. Most of them were taken in Pripyat, the abandoned city inside the Exclusion Zone that was once home to 50,000 people. The entire population was forced to flee within hours of the explosion, leaving everything behind. Pripyat today exists as a sort of Soviet time capsule, a bustling city left in suspended animation, complete with hammers, sickles, and no shortage of radioactive detritus that was once the stuff of regular, everyday lives. Kids’ toys, a ferris wheel, a classroom chalkboard. It’s these everyday items that leave the most lasting impression — a perversion of normalcy that drives home the magnitude of the tragedy. You are free to wander as you please. We had the site almost entirely to ourselves, walking through apartment blocks, kindergarten classrooms, a high school, a hotel.
When the reactor blew, Soviet helicopters dumped sand and clay over the exposed core, and later the building was encased in thousands of tons of concrete — a structure that become known as “the sarcophagus.” In the photo above (an uncropped version appears with the others below), our tour guide aims his dosimeter at the sarcophagus. The reading you see on the machine is about sixty times normal background radiation, and we were allowed to remain here only for about ten minutes.
I should note that reactor four no longer looks like that. In 2016 authorities in Ukraine completed the installation of a mammoth protective dome, concealing the remains within a 25,000-ton shell, made of steel, that looks like a cross between a football stadium and an airship hangar. What you see today is a much more sterile, less jarring aesthetic — though I suppose that’s part of the point.
ALL PHOTOGRAPHS BY PATRICK SMITH
That last one is a shot of downtown Kiev, the pretty Ukrainian capital. A metropolis of almost three million people, Kiev sits only about 80 miles south of Chernobyl. Prevailing winds saved the city from disaster, carrying the fallout in the opposite direction, north into Belarus. From there it diffused across northern Europe.
The items below were souvenirs, I guess you’d have to call them, scavenged from Pripyat. Among them are a 1984 copy of Pravda, the Soviet state newspaper; some vintage postage stamps, and what appears to be a school report card, found inside the Pripyat high school. Perhaps some Ukrainian speakers out there can help translate some of this. I’d love to know more about the report card — names, dates, anything.
The bottom shot is from a roll of exposed film, found on the floor in one of the other buildings.
Hopefully these items haven’t turned my apartment radioactive.
I had been in the Soviet Union about a month before the accident, visiting Moscow and Leningrad (as St. Petersburg still was called). Among the highlights of that trip were my flights aboard Aeroflot. I got to ride a Tupolev Tu-154 from Moscow to Leningrad, and then a Tu-134 from Leningrad to Helsinki. Flights to and from the U.S. were with Finnair. Finnair’s DC-10s, used between New York and Helsinki, featured an absurd, ten-abreast economy section.
Apple juice. I remember the Aeroflot flight attendants serving plastic cups of apple juice.
It dawns on me, too, that my travel habits are at times decidedly macabre. In addition to my trip to Chernobyl, I’ve been to the Auschwitz-Birkenau complex in Poland, and to the various Killing Fields sites around Phnom Penh, in Cambodia.
WILLIAM LANGEWIESCHE, long one of the country’s preeminent journalists, has hit one out of the park with his story in the most recent Atlantic about the disappearance of Malaysia Airlines flight 370. It’s by far the best-written, best-researched, and all-around most compelling piece one can read on the topic. It also has brought me to terms as to what most likely happened on the night of March 8, 2014.
From the beginning I’ve been afraid of the rogue pilot theory — the idea that one of the pilots, presumably Captain Zaharie Shah, was responsible for the flight’s disappearance. I suppose this is partly out of pride. I don’t want it to have been Zaharie, because the idea of the captain hijacking his own aircraft and killing over two-hundred people shames the entire profession.
I’ve instead been noncommittal, usually taking the “two main possibilities” route. After acknowledging at least the chance of Zaharie being the culprit, I propose a second, more complicated scenario in which both pilots became incapacitated:
First, there’s a rapid or explosive cabin decompression. The decompression is caused by an explosion of the crew oxygen bottle, down in the avionics compartment. This not only knocks out the pilots’ oxygen supply, but causes many of the plane’s electronics to fail as well, thus explaining the loss of transponder contact, etc. Zaharie and first officer Hamid commence an initial turn-back toward the airport on the island of Penang (this is the hairpin turn seen on the plots), and enter an initial waypoint or two in the navigation system. They’ve donned their masks, but of course there’s no oxygen flowing, and within seconds they’re unconscious. Minutes later, as the drop-down masks are depleted, so is everyone in the cabin.
After passing Penang, the plane turns right, up the Strait of Malacca and past the northern end of Sumatra — again, in accordance with what’s seen on the satellite plots. It does this because… well, who knows what waypoints may have been in the FMS at that juncture. Maybe, in the throes of losing consciousness, the pilots had typed in nonsense. Heck, there could have been a whole sequence of irrelevant waypoints in the box. Or, is there an arrival procedure or an approach into Penang that might trace a similar outline? Later, the plane defaults to heading mode, and off it goes southwards on a long straight line to oblivion.
Of course, while such a theory explains some of the evidence, it doesn’t explain all of it. For example, it doesn’t account for the plane’s electrical systems later powering up again. Neither can it explain, my thoughts above notwithstanding, why certain course changes occurred precisely where and when they did. The flight path. Langewiesche takes this to the limit:
“It is inconceivable that the known flight path, accompanied by radio and electronic silence, was caused by any combination of system failure and human error,” he writes. “Computer glitch, control-system collapse, squall lines, ice, lightning strike, bird strike, meteorite, volcanic ash, mechanical failure, sensor failure, instrument failure, radio failure, electrical failure, fire, smoke, explosive decompression, cargo explosion, pilot confusion, medical emergency, bomb, war, or act of God — none of these can explain the flight path.”
The problem with my scenario is that it requires further and further layering to make sense. Not just this happened, but also that — and so on.
The only thing that easily and neatly accounts for all of it is Zaharie.
Occam’s Razor. And I think the author has changed my mind, particularly after revealing what other people said and knew about the captain and his personal life — things the Malaysian government has been all too keen on concealing.
Then there’s Zaharie’s home simulator. The idea of a 777 captain owning and spending time on a home simulator in the first place is rather strange. What they found inside it is stranger still. Explains Langewiesche:
“Of all the profiles extracted from the simulator, the one that matched MH370’s path was the only one that Zaharie did not run as a continuous flight — in other words, taking off on the simulator and letting the flight play out, hour after hour, until it reached the destination airport. Instead, he advanced the flight manually in multiple stages, repeatedly jumping the flight forward and subtracting the fuel as necessary until it was gone.”
While this mystery will never be conclusively solved, that’s about the closest thing to a smoking gun we’re liable to find. As Langewiesche puts it, rather understatedly,”The simulator flight cannot easily be dismissed as random coincidence.”
Does locating the wreckage even matter anymore? The voice and data recorders are out there somewhere, nestled invisibly in some immense undersea fissure or canyon, in the ink-black darkness beneath thousands of feet of seawater. The search vessels may have swept right over them. But from the start, they were never likely to tell us much. Malaysian investigators, meanwhile, bumbling and obfuscating from the outset, operating on behalf of a corrupt and secretive regime, want only for the entire story to, if you’ll pardon the irony, vanish forever.
“The important answers probably don’t lie in the ocean but on land, in Malaysia,” Langewiesche concludes. “That should be the focus moving forward. The riddle may not be deep. That is the frustration here. The answers may well lie close at hand, but they are more difficult to retrieve than any black box.”
THE FIRST TIME I saw it was in the fall of 1992, walking along the Revere Beach seawall in the company of our family Weimaraner. It approached from the northeast, head on, lumbering down the coastline. My initial though was Aer Lingus. The afternoon sun had turned blue into green, the forward fuselage taking on the distinctive mossy hue of the Irish national carrier, whose 747s were a regular sight at Logan. But then, as the jet swung closer and into profile, green went blue and I could see, clearly and with some astonishment, that it was Air Force One.
The plane passed less than a thousand feet overhead, then sank past the hills of Beachmont toward runway 22L. I remember it fishtailing slightly — a wobble and a yaw — and silently chuckling. Not even the President’s plane is immune to the push of a good crosswind.
It was a handsome sight. One thing that has always pleased me about Air Force One is the modesty of its livery. Conceived by the famous industrial designer Raymond Loewy during the JFK administration, it’s a look that has gone mostly unchanged for six decades. And for good reason. If you ask me, Air Force One is easily the most elegant state aircraft in the world. The current version, a modified Boeing 747-200 (there are two of them, actually), carries virtually the same markings as the old 707 it superseded: the sweeping forward crown, the Caslon typeface and simple tail hash. The old-timey window stripe and subtle gold highlights, in concert with a couple of judiciously placed flags and the Presidential seal, give the plane a dignified, statesmanlike demeanor. It’s patriotic in the best sense of the word: proud but a little humble.
Then, last summer, Donald Trump announced his intentions to change Air Force One’s livery. He wants to change it because of course he does. Declaring the plane’s robin’s egg blue under-trim a “Jackie Kennedy color,” Trump said he’d prefer something “more American” instead.
Understandably, this made a lot of people nervous. While the paintjob could stand some updating, this is a man whose aesthetic leans heavy on the gold and gaudy — more Saddam Hussein than Jackie Kennedy — and isn’t remotely humble. The resulting scheme was bound to be garish.
Among those who found the idea distressing were U.S. Air Force Brass, countless Americans with good taste, and presidential historian Michael Beschloss. “Why would anyone want to discard an Air Force One design that evokes more than a half-century of American history?” asked Beschloss in Axios magazine. “Every time you see that blue trim and the words ‘UNITED STATES OF AMERICA’ spelled out in the same typeface as an early version of the Declaration of Independence, it brings back JFK landing in Germany to speak at the Berlin Wall, Richard Nixon flying to China, Ronald Reagan stepping off the plane to see Gorbachev in Iceland, and a thousand other scenes of Presidents in our past.”
This week, a new design was revealed. On Wednesday, in an interview with ABC News, Trump held up a poster showing a revised Air Force One, in a livery timed to debut with the delivery of two replacement 747-8 aircraft, on schedule for 2024.
Seeing the headline, I held my breath. I was ready for all manner of over-the-top Trumpian fanfare: star-spangled banners, angry eagle talons, fireworks, maybe a portrait of the Donald himself on the tail. But when I clicked and took a look, I almost couldn’t believe it. It’s completely inoffensive. Boring, but inoffensive.
The fuselage is navy blue across the bottom, with a bold red cheatline riding above. The “United States of America” font appears mostly unchanged. I’m a huge fan of old-fashioned cheatlines — that’s the horizontal striping that runs across the windows — and this one is handsome.
Is it better than the design we already have? No. And if a change was really necessary, any of a dozen other designs would have been better. But it’s respectful, dignified, and it could have been a lot worse. Indeed, one doubts that Trump himself had much or any input, other than to sign off what, most likely, was created by a team somewhere.
The biggest negative is the tail. A jetliner’s tail design is arguably the most critical part of any livery, and here they didn’t even try. There’s nothing there. Just a too-big flag. I recommend a simple tweak to the existing tail instead. Turn the blue hash mark (what do you call that thing?) to red, and there you go. The blue bottom, meanwhile, is too heavy and too rigidly defined. The blue should curve as it meets the forward wing root, similar to the Kennedy hue that’s there today, then reverse-taper beneath the tail. See my rendering, below. I’ve also added a nose swoosh.
Like it or hate it, there’s no guarantee this thing will ever see the light of day. No sooner did Trump show his poster on television when U.S. House of Representative Democrats passed an amendment requiring Congress to approve any changes to Air Force One’s design. Trump himself could well be out of office before the new 747s enter service, and the whole thing could easily be scuttled.
Trump, though, told ABC’s George Stephanopolos that he’s doing this “for other Presidents, not for me.” Maybe, but nonetheless Congress is doing the right thing. This shouldn’t be any President’s call. Air Force One belongs to the nation, not to the President, and its livery shouldn’t be subject to the whims of whomever is holding office at the time. (A Fox News host said, on air, that the amendment was passed by Democrats “because they hate freedom.”)
Trump also says the new 747-8s will be “much bigger” than the current, 747-200 variant. That depends on the definition of “much,” I guess. The 747-8 is 18 feet longer than the -200. The wingspan difference is just under 29 feet, or about fourteen feet per wing, which isn’t a lot when the total span is 225 feet.
Author’s design. Now that’s better.
Officially, “Air Force One” is merely a radio call sign, not the name of a particular aircraft. Any United States Air Force plane with the President on board is Air Force One. Normally this is the 747 we’re familiar with, but occasionally it’s a much smaller 757 or a Gulfstream jet. The President’s helicopter, operated by the U.S. Marines, is “Marine One.”
In 1959, Dwight Eisenhower’s modified Boeing 707 became the first aircraft to use the Air Force One designation. Prior to that, various propeller planes were supplied by the armed forces or contracted commercially for the job. In 1943, Franklin Roosevelt traveled to the Casablanca Conference in a Pan Am flying boat, the Dixie Clipper, celebrating his 61st birthday in the plane’s dining room. Roosevelt himself had created the Presidential Pilot Office to supply the President and his staff with air transportation.
Elsewhere heads of state and their officials do it similarly — or differently, depending. Some travel in standard military transports or will borrow jets from their country’s national airline. Others arrive in stylish airborne limos not unlike our Presidents. For reasons not entirely clear, when Kim Jong-un met with Donald Trump in Singapore in 2017, he arrived from Pyongyang in a chartered Air China 747.
During the 1990s at Logan, I remember, it wasn’t unusual to spot a Saudia Airlines L-1011 TriStar, chocked and secured for the weekend at the north cargo ramp. As the story went, members of the Saudi royal family would drop in for shopping junkets or to visit relatives at local colleges, making use of the huge jetliner the way one might borrow a company car.
FRIDAY, MAY 25th, 1979, was a sunny day in Boston. I was in seventh grade at the time, and a diehard airplane buff who spent pretty much every weekend planespotting from the old observation deck at Logan airport. I was home that afternoon, sitting in the dining room of the house I grew up in, when at a little past four o’clock the phone rang. It was a friend from school. He told me to turn on the television.
It had been a sunny day in Chicago, too, when at 3:04 p.m. local time, American Airlines flight 191, a McDonnell Douglas DC-10 with 271 people on board, roared down runway 32L at O’Hare International Airport, headed for Los Angeles. Just as the plane lifted off, its left engine broke loose. The entire engine, which weighed about eight thousand pounds, together with its connecting pylon and about three feet of the wing’s leading edge, flipped up over the wing and crashed back onto the runway. As the engine ripped away, it severed hydraulic lines, releasing the hydraulic pressure that held the wing’s leading edge slats, deployed to provide critical lift during takeoff, in place. The slats then retracted, causing the left wing to stall. This forced the jet into a violent roll onto its side, beyond 90 degrees, from which it never recovered.
If the pilots had understood what they were dealing with, they could have avoided catastrophe. Simulations would later show that if the crew had reduced power on the right-side engine, they may have been able to counteract that fatal roll to the left. But this ran counter to everything a pilot is taught to do during an engine failure on takeoff. In addition, the engine separation had caused a cascade of electrical malfunctions, knocking out critical flight instruments and alerting systems, including the stall warning and a slat position indicator.
The cockpit voice recorder also had failed. We’ll never know what Captain Walter Lux and his crew said in those final moments, but presumably the only thing they knew for certain is that they’d lost all power in the left engine. It’s doubtful they had any inkling of the separation, the wing damage, the slat retraction or the stall — the reasons for their sudden, sickening loss of control, or how to stop it. There simply wasn’t time.
Only 50 seconds after liftoff, and now banked at 112 degrees, the DC-10 slammed into a field and trailer park less than a mile from the end of the runway, exploding into a gigantic fireball. All 271 passengers and crew were killed, along with two people on the ground. With 273 fatalities, the crash of flight 191 was, and remains, the deadliest air disaster in U.S. history.
Descriptions of the accident are jarring enough. But we can see the horror, too, quite literally. Because a man named Michael Laughlin captured what might be the most haunting aviation photograph ever taken: the stricken jet literally sideways in the sky, only a few seconds before impact, in the throes of that ghastly twist to the left. And then the explosion.
In the aftermath, the National Transportation Safety Board determined that an unseen crack in the support pylon is what caused the engine to break loose. They put most of the blame on American Airlines for faulty practices, including the use of a crane and forklift to support the engine during routine maintenance. They also cited design flaws in both the pylon the wing slat system, as well as a lack of FAA oversight of air carrier maintenance protocols.
On June 6th, after cracks turned up in several other DC-10 engine pylons, the FAA suspended the plane’s operating certificate. For five weeks, no DC-10s, domestic or foreign, were allowed to operate within the United States.
This was just the latest setback for the DC-10, and not even the worst. Five years earlier, in what is still the world’s fourth-deadliest air disaster of all time, a Turkish Airlines DC-10 crashed after takeoff from Orly Airport outside Paris, killing 346 people. The accident was traced to a faulty cargo door design. (The same door had nearly caused the crash of an American Airlines DC-10 two years earlier.) As later revealed in Samme Chittum’s excellent book, “The Flight 981 Disaster,” McDonnell Douglas had hurriedly designed a plane with a door that it knew was defective; then, in the aftermath of Paris, they’d tried to cover the whole thing up. It was reckless, maybe even criminal.
After the grounding in ’79, the DC-10’s reputation had been so shattered that airlines began removing the “McDonnell Douglas DC-10” decals once proudly affixed to their hulls. At American, where it once said “DC-10 Luxury Liner” in red and blue paint near the nose, it now said, “American Airlines Luxury Liner.”
And so it hardly needs saying that, forty years later, the story of the DC-10 reminds us in no small way of the ongoing drama involving Boeing’s 737 MAX. The similarities are uncanny: Multiple catastrophes, a grounding, a lack of regulatory oversight and an aircraft manufacturer accused of negligence and shoddy design.
And in the end, what?
In the case of the DC-10, the airplane soldiered on. Fines were paid, lawsuits were settled, technical fixes were put in place. Sure, there would always be a tiny number of travelers who, out of fear or to make a point, would never again set foot on DC-10. But these were a tiny minority, of no consequence to airlines. For everybody else… they forget, they move on. Time works wonders when it comes to restoring trust.
I suspect the same will hold true for the 737 MAX. The other day I read about a poll in which more than half of all Americans said they didn’t know that a crisis involving the 737 MAX even exists. Other surveys reveal that of those who do know, the vast majority will have no qualms about stepping back on board once the jet’s problems are ironed out. As the headlines have it, it’s doom and gloom for Boeing, its customers, and the MAX itself. But love or hate the 737, I’m not buying it. It might take a while for Boeing, the FAA, and the other vested parties to figure this one out, but I’m guessing the planes will be back soon enough, safely, with their cabins full of travelers.
For most airlines, the grounding of the MAX has been an inconvenience with, for now, minimal financial repercussions. There simply aren’t that many of them in existence. For airlines like American and Southwest, we’re talking about a few dozen aircraft out of several hundred. In ’79, with the DC-10, it was different. DC-10s made up the bulk of the widebody fleets at both American and United. At Continental Airlines, the plane represented a quarter of the airline’s total number of seats. At National Airlines it was almost half.
For a thirteen year-old airplane nut in Boston, the most exciting thing about the grounding was a temporary influx of exotic airplanes into Logan. For a month carriers would substitute other types. United’s DC-10 to Chicago became a 747 — the only 747 I’d ever seen in that carrier’s livery. Swissair brought in DC-8s, Lufthansa sent 707s. And so on. New planes, new colors. I couldn’t get to the airport fast enough.
You can always count on a kid, I guess, to find a silver lining in something so awful as a plane crash.
The O’Hare disaster also takes us back to a time when plane crashes were disturbingly frequent. In the four-year span from 1977 through 1980 there were twenty major air crashes worldwide, killing over 3,400 people. Five of those occurred in 1977, including the Tenerife catastrophe that left 583 people dead. In ’78 there were four accidents, including the PSA midair collision over San Diego and the crash of an Air India 747. In ’79, in addition to flight 191, was the Western Airlines crash in Mexico City and the Air New Zealand sightseeing disaster in Antarctica (both DC-10s, at it happened). A year later we saw six crashes, including the infamous Saudia L-1011 fire that killed 301.
That’s pretty staggering, until you remember that in 1985, twenty-seven major crashes would kill almost 2,500 people. That’s right, twenty-seven crashes in a twelve month span. Among these were the Japan Airlines crash outside Tokyo with 520 fatalities, the Arrow Air disaster in Newfoundland that killed 240 American servicemen, and the Air India bombing over the North Atlantic with 329 dead. Two of history’s ten worst disasters (JAL and Air India) happened within two months of each other!
1985 was an unusually bad year, but from the dawn of the Jet Age all the way into the 1990s, it wasn’t unusual to see five, six, or eight or more major crashes (or bombings) annually, many of them on U.S. soil. Around the time of the Tenerife collision in ’77, I started keeping newspaper clippings. Whenever there was an accident, anywhere in the world, I would snip the related articles from the paper and put them into a shirt box. By the end of junior high, that box was jammed full.
Nowadays, one or two accidents in a year is big news. The number of commercial aircraft worldwide has more than quadrupled since the 1980s, carrying over five times as many passengers. Yet, per passenger-miles flown, flying is an estimated six times safer. Air crashes get a lot of attention — maybe more than they’ve ever gotten — both because and in spite of how infrequently they occur.
YOU’VE DONE A PRETTY GOOD JOB staying clear of the whole social media thing. But then your agent says, in that certain tone of hers, “You really ought to join in.” If only for promotion, she says. It helps sell the book. It helps promote the site. Gets your name out there, and all of that.
And so, late to the party as usual, Ask the Pilot now has an Instagram feed. Is that the right word, feed? Channel, account, timeline… whatever you call it, it’s there, and you’re invited to subscribe.
Most of the photographs will be travel-oriented, though plenty too will be of and from airplanes and airports. Initially, to get things going, I’ll be recycling some older pics from trips and flights around the world. Look for ASKTHEPILOT.
We already went through this with Facebook and Twitter, you might recall.
On Facebook you’ll find three separate pages. The first is the official Ask the Pilot fan page (can it be a fan page if the author himself is creating it?). In addition to the latest posts and stories, it features photos and tidbits that aren’t part of the ATP site itself.
The second page is my personal page. This one is for friends, family, and my more devout and obsessive groupies.
THE CAUSE of the crash of Aeroflot flight 1492 remains unclear. The Sukhoi Superjet-100 had just departed Moscow’s Sheremetyevo Airport, bound for the northern Russian city of Murmansk, when the crew made an emergency return after a lightning strike caused a problem with their communications equipment. The jet then landed hard, bounced and exploded. Forty-one people were killed.
The crash itself may have had nothing to do with the lightning strike. Planes are designed to withstand lightning strikes (see below), and almost never do they cause serious problems. Based on videos, it appears the plane touched down rather violently, and the impact of the landing is what touched off the fire and crash. It’s possible that a strike could have in some way have affected the plane’s instruments or flight controls; or it might have been pilot error.
While we don’t for sure what caused the crash, there’s evidence that the death toll was higher than it might have been, thanks to the selfish actions of a number of passengers. While a deadly fire raged around them, witnesses say people were nonetheless stopping to collect their carry-ons, clogging the aisles and slowing the evacuation. This is just the latest in a string of disturbingly similar incidents. We saw it in in Las Vegas in 2015; in Chicago in 2016; in Toronto in 2005. Among others.
Flight attendants are yelling, “Leave your bags!” but they’re being ignored. People are digging through the bins for their computers and backpacks; here’s a guy coming up the aisle with his roll-aboard. On YouTube you can find selfie videos from idiotic passengers who thought it was cool to film themselves going down the escape slide.
I cannot overemphasize how unsafe this is. Luggage slows people down, impedes access to the aisles and exits, and it turns the escape slides into a deadly slalom. The incident in Moscow is particularly striking, because while most evacuations are precautionary, this one was a full-blown emergency. The airplane was on fire. If that isn’t reason enough to leave your things and get the fuck off the plane as quickly as you can, then heaven help you.
Even in cases where the plane isn’t on fire or about to explode; still, the crew might not be fully certain of what it’s dealing with, and this is never a situation to take lightly. Seconds count, and the goal is to get everybody out as fast as possible. What at first might seem an abundance of caution can quickly turn to terror. Suddenly there’s a mad rush for the exits, but the aisle is clogged with suitcases dropped by panicked passengers. Your computer, your Kindle, your electric toothbrush, your underwear and your Sudoku books — none of those things is worth risking your life over. Not to mention the lives of the passengers behind you, who can’t get to the door because your 26-inch Tumi is in the way.
And a word about those escape slides: although you can’t always see it in videos or photos, the slides are extremely steep. They are not designed with convenience — or fun — in mind. They are designed for no other purpose than to empty a plane of its occupants as rapidly as possible. You’ll be coming down from over two stories high in the case of a widebody jet, at a very rapid clip, with others in front of you and right behind you. Even without bags people are often injured going down the slides. This is expected. Add carry-ons to the mix and somebody is liable to be killed, smacked on the head by your suitcase or baby stroller.
Airlines and regulators are at least partly to blame for this behavior by not better emphasizing the issue during the safety briefing. Of all the gibberish that is crammed into the typical pre-flight demo, one of the most potentially valuable pieces of instruction is often missing: a warning on what to do — or, more accurately, what not to do — in an emergency evacuation. This should be a bold-print, high-emphasis item in any briefing, stated clearly and loudly. Instead we get complicated, twenty-step directions on how to use a life jacket — as if anybody might remember them as they’re jumping into the water. (I could also mention that while neither is likely, a runway evacuation is a lot more likely than a water landing.)
The briefings need to be shorter and more concise, and this needs to be a part of them. It’s hard to say how much it might help to offset some people’s selfish tendencies and general obliviousness, but it wouldn’t hurt.
Leave your bags behind. It all will be returned to you later, no worse for wear. And if, in that rarest of rare cases, it winds up incinerated, you should be happy to have lost it. Lest it have been you in there.
Planes are hit by lightning more frequently than you might expect — an individual jetliner is struck about once every two years, on average — and are built accordingly. The energy does not travel through the cabin electrocuting the passengers; it is discharged overboard through the plane’s aluminum skin, which is an excellent electrical conductor. Composite aircraft are built with a copper mesh beneath the paint layer that acts the same way. Once in a while there’s exterior damage — a superficial entry or exit wound — or minor injury to the plane’s electrical systems, but a strike typically leaves little or no evidence. If indeed lightning touched off a fire aboard Aeroflot 1492, this would be highly unusual.
In 1993, I was captaining a thirty-seven seater when lightning from an embedded cumulonimbus cell got us on the nose. What we felt and heard was little more than a dull flash and a thud. No warning lights flashed, no generators tripped off line. Our conversation went:
“What was that?”
“I don’t know.” [shrug]
“Might have been.”
Mechanics would later find a black smudge on the forward fuselage.
Aeroflot was founded in 1923 and is one of the oldest airlines in the world. In its heydays in the 1970s it was by far the largest carrier on earth — bigger than all of the U.S. majors combined. After the fall of the Soviet Union it was broken up into dozens of airlines, the largest of which still carries the Aeroflot name.
OVER THE PAST dozen years or so, the trend in airline livery design has been one big race to the bottom. Whenever a carrier announces a change is in the works, we prepare to be disappointed.
And so it went after getting word that United Airlines, the nation’s number two carrier, was about to unveil a new look. A part of me held out hope. Maybe, I thought, they’d bring back a version of the iconic, sorely missed “U” logo that graced the carrier’s tails for so many years. Maybe we’d see something akin to the carrier’s handsome mid-90s look, with the gray top and elegant understriping.
That was always a longshot, of course.
The livery being replaced, which came about after the merger with Continental Airlines in 2010, is itself nothing special. It’s an amalgamation that blended the United typeface with the Continental globe. Bland and ultra-corporate, it looks like something you’d see in a PowerPoint slide. There was plenty of opportunity here for improvement, which brings us to the most important question: is the new livery better?
The answer is no, it’s not. They’ve stayed with the 2010 template; except, now, they’ve sucked away whatever dignity it had.
We start with the “United” title, which has gone big. Big for big’s sake, in a way that looks unbalanced and garish. It’s oddly spaced as well. And, needless to say, you can’t have a livery these days without some annoying “in-motion” theme. United obliges with a mandatory curvy thing — it looks like a garden hose — along the lower fuselage. It’s capricious, worm-like, and worst of all it’s black — black!
The gold accenting is gone from the tail now, too, turning the Continental globe into a sort of fluorescent hub cap. Is this the airline’s excuse for a logo? Do they even have a logo? It’s a gaudy tail in a two-tone blue that’s at once syrupy and overexposed.
Granted this isn’t as terrible as what American Airlines did a few years ago. It’s bold, I’ll give you that, and you can marvel in the simplicity of it. Or, you can call it what it is: boring, and totally without spirit. Like too many identities these days, it’s watered-down, cowardly, half-assed scheme that looks like it was designed in about fifteen minutes.
With both AA and UA having let us down, together with JetBlue’s lackluster scheme and Southwest’s carnival car abomination, this leaves Delta, for now, head and shoulders above the other U.S. majors.
DID YOU HEAR the one about the flight that went to Edinburgh instead of Dusseldorf?
On March 25th, a British Airways regional jet took off from London City Airport (LCY) bound for Dusseldorf, Germany. An hour or so later, the passengers found themselves in Edinburgh, Scotland. Not until the jet touched down did they realize they were headed to the wrong city.
Reporting on this incident has been inconsistent at best. Obviously certain outlets find the whole “the pilots flew to the wrong airport!” angle irresistible. But that isn’t what happened. The pilots, the flight attendants, and their aircraft were dispatched, flight-planned, and fueled for a trip to Edinburgh. That’s where they were expected to fly, and that’s where they went.
The passengers, on the other hand, thought they were going to Dusseldorf. In other words, the plane didn’t fly to the wrong city. The passengers were put on the wrong plane.
The breakdown was apparently between the technical staff (pilots, dispatchers and flight attendants) and the airport personnel who processed and boarded the passengers. The former were Edinburgh-bound. The latter were handling a flight to Germany — or so they thought.
But, you’re thinking, wouldn’t the “welcome aboard” spiel by either the pilots or cabin crew have given things away? We always mention the destination, don’t we? And surely flight attendants would have noticed “Dusseldorf” on the customers’ boarding passes, right? The signs and announcements at the gate, too, would have referenced Edinburgh. Could all of this have somehow been missed?
I, for one, can see it happening. Those PAs are sometimes perfunctory, and how many people are listening to begin with? And I can easily — easily — imagine a scenario where the gate personnel and flight crew found themselves on separate pages, regardless of any boarding announcements or signs above the podium. The fact this was a contract company flying on British Airways’ behalf, rather than an actual BA aircraft operated and overseen by BA personnel, could also have been a factor.
Embarrassing, sure. But not the same thing as two pilots taking their plane to the wrong destination. Which, I’m the first to admit, has happened. More than once…
In 2013, a Southwest 737 destined for Branson, Missouri, instead ended up at a small general aviation field nearby, touching down on a runway less than four thousand feet long. Only a few months earlier, a 747 freighter operated by Atlas Air found itself at the wrong airport in Kansas. In June, 2004, a Northwest Airlines flight from Minneapolis to Rapid City mistakenly landed at Ellsworth Air Force Base. And, in 1995, a Northwest DC-10 touched down in Brussels instead of Frankfurt.
The idea of highly trained aircrews with troves of technology at their behest landing astray sounds, I’ll agree, amusing, quaint, or even patently ridiculous. So how does it happen? Improperly keyed coordinates? A navigational computer gone crazy? Or is there a more visceral, seat-of-the-pants explanation, such as a tired crew mixing up a pair of similar-looking runways.
If there’s a common thread, it’s that often in these cases pilots were flying what we call a “visual approach.” Most of the time, jetliners land using what we call an ILS (instrument landing system) in which controllers guide us onto a pair of radio beams — one vertical, the other horizontal — that form a sort of crosshair that we track to the runway, either manually or by coupling the ILS to the plane’s autoflight system. There are also what we call “non-precision” instrument patterns, in which a GPS-guided course takes you to a few hundred feet or less above the pavement. But a visual approach, as the name implies, is almost entirely pilot-guided. This is when high-tech goes low-tech. Essentially you eyeball the airport through the windshield, report “field in sight” to ATC, get your clearance and go ahead and land. When available we back-up a visual with whatever electronic landing aids might also be available (an ILS signal, usually), But sometimes our orientation is based entirely on what we see through the window.
Visual approaches are very common, very routine, even at the biggest and busiest airports. Airline pilots perform thousands of these approaches every day without incident. However, you need to be sure of what you’re looking at. This is particularly true at smaller airports, where cues on the ground (roads, coastlines, buildings etc.) aren’t as obvious, and where you might not be familiar with the surroundings. Certain combinations of circumstances make a mistake more likely: a nighttime visual with low-altitude maneuvering, for instance, at an airport you’re not used to, with perhaps a similarly laid-out airport nearby.
In the winter of 1990 I was copiloting a small turboprop when we were cleared for a nighttime visual approach to New Haven, Connecticut. As it happens, the lights and orientation of nearby Bridgeport, Connecticut, appear strikingly similar to those of New Haven. After a minute or two we realized the error and corrected course. All of this happened far from the runway, and, somewhat in our defense, we were flying a 15-seat Beech-99 — a vintage relic from the Age of Aquarius, with as many electronic accoutrements as my mountain bike.
In addition to whatever human errors catalyze such events, weather and air traffic control (ATC), to name two, can lend a hand in getting from point A to, as it were, point C. In the case of that DC-10 finding its way to Belgium instead of Germany, air traffic controllers had been given the wrong information, and began issuing a long and complicated series of vectors and course changes to the crew, sending it toward Brussels. Airspace in Europe is complex and congested, and roundabout routings aren’t uncommon. Thus it wasn’t necessarily obvious to the crew that they were being led astray. By the time the crew realized they were being vectored to Brussels, they decided to land there rather than have to recalculate fuel reserves and orchestrate a last-second re-routing. In the end, it was probably safer to land at the wrong airport than the right one.
The act of landing a plane is, on one level, inordinately simple. At the same time, it’s a maneuver beholden to technology. It can be one, or the other, or both, depending on circumstances: the aircraft type, the approach being flown, the weather, the airport. There are times when an old-school, seat-of-the-pants skill set is exactly what’s required. Other times it’s all but impossible to find a runway without help from the instruments and screens in front of us. Either way, though, it comes down to judgment and decision-making. Despite everything you hear about autopilot and the alleged sophistication of modern jetliners, it’s the fight crew — the pilots — who are very much flying the airplane. Sometimes it takes an embarrassing mistake to make this point clear.
Ethiopian Airlines CEO Tewolde Gebremariam at the 737 crash site.
UPDATE: May 20, 2019
The 737 MAX saga is spinning in so many different directions now that I’ve decided to lay off the updates for a while. The media, both big and small, have been doing a surprisingly good job with this story — and all of its various substories — and at this point there’s little that I can add, from the perspective of an airline pilot or insider, that I haven’t already said in the installments below.
UPDATE: April 13, 2019
What a mess. Boeing is getting knocked around by everyone from members of Congress to late-night comedians. The MAX’s certification program is under scrutiny, airlines are canceling orders, and passengers everywhere are scared. The FAA is facing accusations that it took far too long to order the MAX’s grounding (after numerous other countries had already done so), and that it basically permitted Boeing to self-certify an unsafe aircraft.
We keep hearing, too, about what a horrible black mark this is not merely against Boeing, but against American aviation’s place in the world. We are no longer the global leader in air safety, no longer the “gold standard,” whatever that means exactly, as several articles have described it. This is maybe just another example of the weird phenomenon known as American exceptionalism, but each time I hear it, I keep going back to the DC-10 fiasco in the 1970s.
In 1974, in one of the most horrific air disasters of all time, a THY (Turkish Airlines) DC-10 crashed after takeoff from Orly Airport outside Paris, killing 346 people. The accident was traced to a faulty cargo door design. (The same door had nearly caused the crash of an American Airlines DC-10 two years earlier.) McDonnell Douglas had hurriedly designed a plane with a door that it knew was defective; then, in the aftermath of Paris, tried to cover the whole thing up. It was reckless, even criminal. Then, in 1979, American flight 191, also a DC-10, went down at Chicago-O’Hare, killing 273 — to this day the deadliest air crash ever on U.S. soil — after an engine detached on takeoff. Investigators blamed improper maintenance procedures (including use of a forklift to raise the engine and its pylon), and then found pylon cracks in at least six other DC-10s, causing the entire fleet to be grounded for 37 days. The NTSB also cited design flaws in the engine pylon and wing slats, quality control problems at McDonnell Douglas, and “deficiencies in the surveillance and reporting procedures of the FAA.”
That’s two of history’s ten deadliest air crashes, complete with design defects, a cover-up, and 619 dead people. And don’t forget the 737 itself has a checkered past, going back to the rudder problems that caused the crash of USAir flight 427 in 1994 (and likely the crash of United flight 585 in 1991). Yet the DC-10, the 737, and America’s aviation prestige along with them, have persevered. If we survived the those scandals we can probably manage this. I have a feeling that a year from now this saga will be mostly forgotten. Boeing and its stock price will recover, the MAX will be up and flying again, and on and on we go. This is how it happens.
There’s also a lot being made of the FAA’s more or less outsourcing aircraft certification to Boeing. This is frustrating, and ironic, because air travel has never been safer, and it’s partly because, not in spite of, the close relationship and collaborative efforts between regulators, airlines, manufacturers, pilot groups, and so on. (A good example is the self-reporting program between pilots and FAA, which has been very successful and has kept dangerous trends from being driven underground.) Bear in mind how much these parties stand to lose should a tragedy occur. A crash can destroy an airline outright. It’s in the interest of all these entities to play things as safely as possible.
Did something go wrong in the 737 program? Are Boeing and the FAA jointly responsible? Probably. But I don’t believe anybody was intentionally reckless. That’s an important distinction, and for the most part the relationships between industry and regulators has been a productive one. You can’t say that about banking, perhaps, but in aviation it seems to work. The remarkable safety record we’ve enjoyed over the past twenty years bears that out, absolutely.
For the airline passenger, these can seem like scary times. Air crashes, perhaps more than any other type of catastrophe, have a way of haunting the public’s consciousness, particularly when the causes are mysterious. My best advice, maybe, is to turn off the news, take a step back, and try to look at this through a wider lens. The fact is, Lion Air and Ethiopian notwithstanding, air travel has never been safer than it is today. Two fatal crashes in five months is tragic, but in decades past it wasn’t unusual to see ten, fifteen, or even twenty air disasters worldwide in a given year. Nowadays, two or more is downright unusual. Here in the United States there hasn’t been a large-scale fatal crash involving a mainline carrier in nearly twenty years — an absolutely astonishing statistic. There are far more planes, carrying far more passengers, than ever before, yet the accident rate is a fraction of what it once was.
I have a question, however:
One of the things I can’t help wondering about is why the 737 MAX needed to exist in the first damn place. Somewhere deep down, perhaps the heart of this whole fiasco is stubbornness — that is, Boeing’s determination to keep the 737 line going, variant after variant, seemingly forever. Instead of starting from scratch with a new airframe, they took what was essentially conceived as a regional jet in the mid-1960s, and have pushed and pushed and pushed the thing — bigger and bigger engines, fancier avionics and more seats — into roles it was never intended for.
For a jet of its size, it uses huge amounts of runway and has startlingly high takeoff and landing speeds. The passenger cabin is skinny and uncomfortable; the cockpit is incredibly cramped and noisy. I don’t care how many changes and updates the plane has undergone; at heart, it’s still a blasted 737 — a fifty year-old design trying to pass itself off as a modern jetliner. The “Frankenplane,” I call it. Look at that tell-tale nose and windscreen. Do you recognize that? It’s the 707, from 1958, unchanged.
I’m not saying this is the reason, directly, for what happened in Indonesia or Ethiopia, but is it maybe not time, at last, to move on from the 737 platform?
UPDATE: April 6, 2019
This just makes you shake your head.
What seems to be the case, based on analysis of the voice and data recorders from the doomed Ethiopian Airlines flight 302, is that the pilots did, as they should have, engage the plane’s pitch trim disconnect switches in a frantic attempt to regain control after a malfunctioning MCAS system forced the plane’s nose toward the ground. This pair of switches, on the center console near the thrust levers, killed power to the entire automatic pitch trim system, including MCAS, and should have allowed the pilots to maintain a normal flightpath using manual trim and elevator. Manual trim is applied by turning a large wheel mounted to the side of that same center console. Elevator is controlled by moving the control column forward or aft.
Yet they did not, could not, regain control. The reason, many now believe, is a design quirk of the 737 — an idiosyncrasy that reveals itself in only the rarest of circumstances, and that few 737 pilots are aware of. When the plane’s stabilizers are acting to push the nose down, and the control column is simultaneously pulled aft, a sort of aerodynamic lockout forms: airflow forces on the stabilizers effectively paralyze them, making them impossible to move manually.
Aboard flight 302, the scenario goes like this: Commands of the faulty MCAS are causing the automatic trim system to push the nose down. The pilots, trying to arrest this descent, are pulling aft on the control column. Thus setting up this scenario perfectly. The trim forces are stronger than the control column forces, which is why pulling back on the column has no effect. But now, with power to the trim system shut off, they can lift the nose by manually by rotating the trim wheel aft, relieving that unwanted nose-down push. But the wheel won’t move. Believing the manual trim is itself broken, the pilots then reengaged the auto-trim. MCAS then kicks in again, pushing the nose down even further. What’s worse, as the plane’s speed increases, the lockout effect intensifies. And so with every passing second it becomes more and more difficult to recover.
The correct course of action would be to relax pressure on the control column, perhaps to the point of pushing the nose down even further. This will free the stabilizers of the aerodynamic weirdness that is paralyzing them, and allow the trim wheel to move, realigning the stabilizers to a proper and safe position. For the pilots, though, such a move would be completely counterintuitive. Instead, they do what any pilots would be expected to do under the circumstances. Turns out it’s the wrong thing, but really they have no way of knowing.
It’s possible, or probable, that the pilots of Lion Air flight 610 faced exactly the same situation, with the same result.
Apparently, pilots of older-generation 737s — long before there was MCAS — were aware of the lockout potential, and some were trained accordingly. (I flew the “classic” 737-200, briefly, about twenty years ago, but have no memory of it one way or the other.) However, as an obscure phenomenon that no pilot was likely to ever encounter, it was eventually forgotten as the 737 line evolved, to the point where no mention of it appears in the manuals of later variants.
Circles, left to right:
1. Electric trim switches. With the autopilot engaged the pitch trim system operates automatically. With the autopilot off, the pilot controls the trim by manipulating these switches forward or aft, usually with his or her thumb.
2. Trim wheel. This is the wheel that the pilot will rotate forward or aft to control trim manually.
3. The disconnect switches. These kill power to the trim system. Auto-trim and the thumb switches are now shut off; trim is adjusted using the wheel in the second circle.
UPDATE: March 29, 2019
ON MARCH 10th, Ethiopian Airlines flight ET302, a Boeing 737 MAX bound for Nairobi, crashed after takeoff from Addis Ababa, killing 157 people from more than thirty countries. Five months earlier, 189 people perished after Lion Air flight JT610 went down near Jakarta, Indonesia, under eerily similar circumstances. Both planes were brand new 737 MAX jets. Both crashed shortly after takeoff following a loss of control.
Although findings from the voice and data recorders pulled from the Ethiopian wreckage haven’t been released yet, it’s all but assumed that flight succumbed to the same flight control malfunction that brought down Lion Air. The 737 MAX has a deadly design problem, and Boeing needs to fix it. In the meantime, all MAX jets remain grounded worldwide.
The culprit is something called MCAS, which stands for Maneuvering Characteristics Augmentation System, a system that adjusts control feel as the plane’s nose pitches upward, effectively nudging it downward.
MCAS operates in the background, transparently and automatically — there’s no on or off switch, per se — and only during a very narrow window of the jet’s flight envelope. This is not something that occurs in normal, day-to-day operation, but certification requires it for those occasions when, for whatever reason, the plane reaches unusually steep climb angles. To raise a plane’s nose, the pilot pulls back on the control column. As the nose pitches further and further upward, the control forces required to maintain this action are supposed to become heavier. This helps keep pilots, and/or the autopilot, from inadvertently stalling the plane — that is, exceeding what we call the “critical angle of attack,” at which point the wings run out of lift and the plane ceases to fly. On the 737 MAX, however, certain aerodynamic factors, including the placement of its very powerful engines, result in control forces actually becoming lighter as it approaches the point of stall. Because of this the plane would not meet certification standards. And so MCAS was engineered in to properly adjust the feel.
Thus there’s a certain beauty to MCAS — provided it works correctly. What’s happening, apparently, is that faulty data is being fed to MCAS by the plane’s angle of attack indicator — a small, wedge-shaped sensor near the plane’s nose that helps warn pilots of an encroaching aerodynamic stall. An impending stall is sensed when there isn’t one, triggering the plane’s stabilizer trim — stabilizers are the wing-like horizontal surfaces beneath the tail — to force the nose down. This sets up a battle of sorts between the pilots and the trim system until the plane becomes uncontrollable and crashes.
What leaves us stymied, though, is the fact that any MCAS commands, faulty or not, can be overridden quickly through a pair of disconnect switches. Why the Lion Air pilots failed to engage these switches is unclear, but unaware of the system’s defect in the first place, we can easily envision a scenario in which they became overwhelmed, unable to figure out in time what the plane was doing and how to correct it. From that point forward, however, things were different. “Though it appears there’s a design flaw that Boeing will need to fix as soon as possible,” I wrote in November,“passengers can take comfort in knowing that every MAX pilot is now acutely aware of this potential problem, and is prepared deal with it.”
Or so it seemed. With the Lion Air crash fresh on any MAX pilot’s mind, why did the Ethiopian pilots not immediately disconnect the trim system? Did a disconnect somehow not work? Was the crew so inundated by a cascade of alarms, warnings, and erratic aircraft behavior that they failed to recognize what was happening? Or, was the problem something else completely? This is the most perplexing part of this whole unfolding drama.
While we wait for the black box results, Boeing this week revealed a suite of hardware and software tweaks that it claims will rectify the issue. This includes incorporation of a second angle of attack indicator, and an alerting system to warn pilots of a disagreement between the two.
Ethiopian Airlines 737 MAX
The largest MAX operators in the U.S. are American Airlines, Southwest and United. Other customers include Alaska Airlines, Air China, Norwegian, FlyDubai, China Eastern and China Southern. The type is most easily recognize by its 787-style scalloped engine nacelles, which earlier 737s do not have.
Founded in 1945, Ethiopian Airlines is the largest carrier in Africa. Westerners hear “Ethiopia” and tend to make certain, unfortunate associations, but this is company with a proud history and a very good safety record. It flies a state-of-the art fleet, including the Boeing 787 and A350, on routes across four continents. Its training department, the Ethiopian Airlines Aviation Academy, has been training pilots for 55 years. Ethiopian’s pilots are distinguished by their handsome, olive green uniforms.
The captain of the doomed flight ET302, Yared Getachew, was a graduate of the highly competitive Ethiopian Airlines Aviation Academy, and had more than 8,000 flight hours — a respectable total. “Yared was a great person and a great pilot. Well prepared,” a former Ethiopian Airlines training captain told me.
The first officer, on the other hand, had a mere two-hundred hours. Airline training is intensive, and as I’ve written in the past, the raw number of hours in a pilot’s logbook isn’t always a good indicator of skill or talent. Nonetheless, if indeed that number is correct (it’s unclear if this refers to his total flight time, or his number of hours in the 737 MAX), that’s pretty astounding. By comparison, the typical new-hire at a U.S. major carrier has somewhere on the order of 5,000 hours. Whether the first officer’s lack of experience had anything to do with the accident, however, is another matter.
Cockpit photo by Vedant Agarwal, New York Times
Thumbnail photo by Michael Toweled, AFP