My family and I were in Washington D. C. Reagan airport waiting for the Air Canda 8823 on 14 March 2023 (the aircraft that was supposed to fly us back home to Montreal).

I was following the flight on Flightradar24. Three hundred feet before touching down the plane executed a go around and diverted to Toronto. The airline blamed high winds even though every other plane seemed to be able to land. Why did they divert to Toronto instead of a nearby alternate and why not back to Montreal?

  • 3
    $\begingroup$ Yesterday's METAR confirms that it was very windy during the time they arrived ("KDCA 141952Z 31021G33KT ..." and "KDCA 142052Z 30019G40KT ..."), but other than that, I'm afraid we could only speculate... $\endgroup$
    – Bianfable
    Commented Mar 15, 2023 at 10:42
  • 3
    $\begingroup$ I don't have a definite answer, I suspect its due to immigration considerations. $\endgroup$
    – GdD
    Commented Mar 15, 2023 at 10:50
  • $\begingroup$ It is ultimately the captain's judgement; maybe the winds were quite marginal and your flight happened to have the most cautious captain of the lot. $\endgroup$ Commented Mar 15, 2023 at 21:05

1 Answer 1


Although this may not fully answer the question, I post this information as such because it looks into the cause (high wind) provided as a reason by the airline, and it is too much for a comment. If any user want to reuse this text as a basis for a more elaborate answer, please feel free to do so.

The aircraft initiated the go-around at approximately 20:15 UTC, from a height of 300 ft.

Other aircraft were able to land in the 30 minutes before and after the occurence. Most did on the first attempt, a few needed a second attempt.

The primary landing runway at the time was RWY 01, few aircraft circled to land at runway 33.

The weather according to the METARs around that time:

METAR KDCA 141952Z 31021G33KT 10SM BKN060 06/M09 A2990 RMK AO2 PK WND 28035/1921 SLP123 T00561089=

METAR KDCA 142052Z 30019G40KT 10SM BKN055 06/M09 A2991 RMK AO2 PK WND 28040/2043 SLP128 T00561089 53006=

The wind was from a direction of 300 to 310 degrees at around 20 knots, with gusts up to 40 knots.

Peak winds (measured during the past hour) were at 35 knots at 19:21 and 40 knots at 20:43 UTZ. These peak winds were from a more westerly direction of 280 degrees.

A gust from 280 degrees at 40 knots would be a 40kt crosswind component on RWY 01. The same gust would be a 26 knots crosswind component on RWY 33.

The demonstrated crosswind component on the Embraer 170 is 28 kts sustained, with gusts to 38 kts on dry runways. On wet runways, the demonstrated crosswind component is 28 kts.

The conditions during the time of final approach were close to the maximum demonstrated crosswind, but probably not exceeding it.

Dulles airport (IAD) has a runway 30, with almost no crosswind component at the time.

I will not speculate to the reasons why no second approach was attempted, and why a diversion back to Canada was decided.

  • 2
    $\begingroup$ They always put in the statement "not considered limiting" on x wind limitations so it remains the Capt's judgement call, and some will go for it and some won't. $\endgroup$
    – John K
    Commented Mar 15, 2023 at 15:48
  • 4
    $\begingroup$ @JohnK indeed. It is a maximum demonstrated value, not a limit. So as the captain, at your own discretion, you can decide to attempt a landing in a stronger crosswind. But it is not easy to explain in front of the investigation committee why you, as the captain that crashed the aircraft, thought you were better at cross wind landings than the test pilots. $\endgroup$
    – DeltaLima
    Commented Mar 15, 2023 at 15:53
  • 2
    $\begingroup$ @JohnK While the manufacturer's AFM manual landing crosswind is demonstrated and not limiting, company policy typically dictates a limit. I am pretty sure AC's policy is limiting. Company policy may also further limit based on FO flying and low-time pilots. Autoland is also a AFM limit, and significantly CAT II and CAT III operations have a 15 kt limit set by the FAA (OpSpec C060) due to the design of ILS. $\endgroup$
    – user71659
    Commented Mar 15, 2023 at 18:40

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .