CRJ900 Overturns on Canada’s Busiest Runway
On the 17th of February 2025, Delta flight 4819 from Minneapolis, Minnesota to Toronto, Ontario suffered a hard landing and flipped over, bursting into flames. The aircraft was fully loaded with 76 passengers and four crew on board.
One passenger wrote about the experience on Reddit:
I felt what seemed like too violent of jerking motions from the wind immediately before landing. We touched down HARD, then bounced back up and started to skid. I’m not sure exactly how we rolled or when, but I was over the wings and could see the orange glow from the fire through the right side windows. We skidded for what seemed like a short time. There was a lot of screaming, though I remember thinking it sounded like we were underwater. Everything was dulled. When we stopped, there was a brief moment of silence before everyone started trying to release themselves. Those of us who could get ourselves down did, and then helped others.
The following is CCTV video of the crash as posted on X:
New surveillance video captures the moment a Delta Airlines jet crashed at Toronto Pearson International Airport, with 75 passengers on board leaving 18 injured. pic.twitter.com/jEDMlpUCqq
— Mr Commonsense (@fopminui) February 18, 2025
This is the audio file from ATCLive at the point of the incident: skip ahead to the two minute mark for the response to the crash.
Audio PlayerAnd VASAviation has published a streamlined account of the key interactions.
The aircraft was a Bombardier CRJ900LR, registered in the US as N932XJ, a long-range CRJ900 popular with regional jets for short- to medium-haul flights. Delta’s CRJ900LR aircraft are configured for 76 seats with six exits (two at the front and four overwing). The aircraft was operated by Endeavor Air, a subsidiary of Delta. They departed Minneapolis-Saint Paul International Airport in Minnesota, US for the 86-minute flight to Toronto.
Here is the METAR at Toronto as they approached:
CYYZ 171900Z 27028G35KT 6SM R24L/3000VP6000FT/U BLSN BKN034 M09/M14 A2993 RMK CU6 SLP149=
METeorological Aerodrome Report from Toronto Pearson (CYYZ) at 19:00Z:
Winds from 270° at 28 knots, gusting to 35 knots
6 statute miles visibility
Runway 24L visual range varying from 3,000 feet to over 6,000 feet, with an upward trend
Blowing snow
Broken clouds at 3,400 feet
Temperature -9°C, dewpoint -14°C
Altimeter 29.93 inHg
Remarks: Cumulus clouds 6/8 coverage, sea level pressure 1014.9 hPa
Delta released a statement that the Endeavor Air flight crew of flight 4819 were experienced and trained in wintry conditions. Rather frustratingly, yesterday they had to release a further statement to confirm that the captain and/or first officer were qualified, countering malicious rumours that the flight crew had failed training events or that they had failed to gain pilot positions at Delta.
Whether or not there were issues with CRM or in the approach, there is no evidence at this time that either pilot had struggled through their training or were anything less than qualified for their positions.
What seems clear is that the aircraft came in to runway 23 fast and flat, as can be seen on this video taken by a pilot in an aircraft waiting on the taxi-way:
New footage of the Delta CRJ-900 crash that occurred at Toronto Pearson International Airportpic.twitter.com/2ilJwqRDDs
— Massimo (@Rainmaker1973) February 18, 2025
The aircraft touched down hard and the landing gear collapsed or retracted. The wing tip struck the runway after which the wing and the tail separated and the fuel from the wing caught fire. The aircraft tumbled and came to a halt upside-down at the intersection of runway 23 and runway 15.
Twenty-one passengers were injured in the crash, with three in critical condition. All passengers have now been discharged from hospital and expected to make full recoveries.
TMZ has posted some further angles of the landing and crash:
From the videos, it appears that the aircraft landed hard without a flare. The wing tip then struck the ground, at which point the wing and the tail section broke. As a result, the aircraft then rolled and came to a halt in a ditch.
It is not yet clear how hard the impact was at landing or why the landing gear collapsed or retracted. Although a wing strike is never a good thing, it is not expected for the wing to rip off like that.
The Canadian TSB are investigating with support from the NTSB.
They arrived at the scene quickly to collect information before the snow changed. The recorders were recovered from the wreckage and sent to the TSB Engineering Laboratory in Ottawa, who confirmed that the data has been downloaded and being analysed. The cockpit recordings during the final approach and the flight data on impact are sure to be of particular interest.
I recall from my IFR days that visibility measurements could vary, but ISTM that reporting 6 miles generally versus .6-1.2 miles on the runway is extreme. Maybe fallen snow was being stirred up some feet above ground level by the wind?
I wonder whether the flat approach was an attempt to get wheels on the ground before the crosswind become unmanageable; whether they were slipping or crabbing, they would have had to do more of it to keep the plane on line if they were bleeding off speed in a flare — and a smaller plane like that would have had a lower landing speed (hence more affected by the crosswind) to begin with.
Kudos to the cabin crew for getting everyone out quickly and mostly safely; I know they practice for many circumstances, but I doubt the plane being upside down is one of them.
From the description of the landing as having increased wind noise just before touchdown, I wonder if the complete lack of flare was due to a change in wind or wind shear just before landing? Almost looks like a carrier approach in the vids; I’d be surprised if they was deliberate.
I was puzzled at first as to what made the aircraft turn over. However, shortly after touch down both wings are still producing substantial lift, and if one wing detaches, the force on the other wing pulls on it like a handle and rotates the aircraft.
There’s a conspiracy theory about the aircraft having “explosive bolts” to detach the wings; if that were true (it’s not), this accident would’ve been the right time to use them and prevent the upset.
Not really much to comment on. Mendel quite correctly mentioned that, with the right wing detached, the lift still produced by the left one would have pulled the aircraft around exactly as happened. But explosive bolts to blow a wing off? in a passenger aircraft? What did you smoke in your pipe, Mendel?
What caused the crash is of course a matter for the official investigation board. But it seems that the runway conditions were decent over the full length and the wind, though strong and gusty, was within the limits. Looking at videos that have circulated widely, the aircraft appeared to be on a stable glide path. But there does not seem to be any attempt to arrest the descent, no flare. The aircraft would have a rate of descent of about 750 feet per minute, too fast. The landing gear would have been able to absorb a hard landing at about 500 fpm. If drifting snow were to be blamed for the absence of a flare, why not react to the call-outs from the radio altimeter> Like “two hundred, one hundred, fifty”. Judging from the video, taken from a distance, the aircraft was landed like a fighter jet onto the deck of an aircraft carrier. It is impossible to see if the pilot “kicked off” a crab prior to a crosswind landing. Nor does the video show any last-moment destabilization or last-minute deviation from the glide path due to wind gusts.
So, awaiting official findings, the only explanation seems to be: Neither the weather nor the runway conditions were outside the limits. The aircraft seemed to be on a stable glide path, failed to flare (reason unknown), resulting in a very had landing.
The aircraft sits low on the ground, maybe the right wingtip hit the ground, maybe the right landing gear was damaged. The wing briefly scraped along the ground, then became detached. Spilled fuel ignited but fortunately fire did not engulf the fuselage. The lift still generated by the left wing forced the entire aircraft to come to a stop inverted, with the right wing behind it.
The cabin crew would not have been trained to handle this kind of scenario. They did a magnificent job and should be rewarded. Truly heroic stuff!
A not really related comment: After watching the videos on the “Fear of ,,,” blog I am suddenly confronted with messages from “X” regarding my access code.
I have never used Twitter and most certainly don’t want to use neo-Nazi Musk’s Russian propaganda machine. (I won’t buy a Swasticar, aka Tesla, either).
How can I get out of X-s clutches, anyone?
If it’s email, mark it as spam? I deserted twittter after the Nazi takeover (although it was a useful fact checker… want to know if something’s a lie? See if Musk or Turnip have said it!).
These messages are probably phishing: they do not originate with Xitter, but with scammers. They aim to have unsuspecting users give them their access codes, and then take over their accounts.
These scammers could have gotten your email address from any badly secured website. You watching videos here on this blog is simply a coincidence.
Agreeing with the others: there’s no way for them to get your contact details by watching a video here and it’s annoying but probably not X/Twitter but scammers sending the same message to hundreds of thousands of people, hoping that some will click through and give away their details. I do try to find other sources for videos when I can, but unfortunately X/Twitter still seems to be very central to aviation reporting.
The NYTimes quotes a preliminary report that the descent rate was ~4x normal and ~2x what the landing gear was rated for; the right main gear collapsed, leading to what we see in the video. https://www.nytimes.com/2025/03/20/world/canada/delta-plane-crash-toronto-report.html
Link to the investigation at the TSB, where you can find the report:
https://www.tsb.gc.ca/eng/enquetes-investigations/aviation/2025/a25o0021/a25o0021.html
The first officer was flying, reduced thrust in response to a gust, and the aircraft lost energy as a result. The captain did not call for or execute a go-around when the GPWS “sink rate” warning sounded (1100 fpm at 50 feet AGL).
As I understand, it is still not quite clear what the exact accident sequence was with respect as to when and why the right main gear and main wing spar failed.