September 05, 2022

Was There A Weather Connection with the Whidbey Float Plane Crash?

A tragedy occur yesterday (Sunday, September 4) with the crash of a float plane over Mutiny Bay of southern Whidbey Island just after 3 PM (see map below, the star indicates the accident site).  

The aircraft was en route from Friday Harbor to Renton.  Like all aircraft accidents, it will be examined comprehensively by the National Traffic Safety Board (NTSB), which is very valuable.  Flying is extraordinarily safe because the causes of past accidents are painstakingly investigated and the system is improved as a result.

Several folks have asked me whether there was a meteorological contributor to this terrible occurrence, and after examining the data, I did find something significant.

Let me show you now.

During the exact time of the accident a front was moving through the regon, and this passage, interacting with local terrain, produced a very sharp low-level wind shift in the lee of the Olympics.  

Here are the surface observations at 3 PM.  To the south of the accident site the winds were from the south (the barbs face the direction from which the wind is coming) while to the north the winds were from the northwest.  


Temperatures were much warmer in the southerly flow compared to the northerly flow.

Such convergence of low-level airflow often accompanies frontal passage over Puget Sound and often produces a band of clouds and precipitation:  the infamous Puget Sound Convergence Zone.

This transition can be very sharp, as a cool wedge of cold, northerly air undercuts and pushes aside warmer southerly flow (see schematic vertical cross section below).


To see how sharp the transition was, let me show you the observations from a surface observing location (Bush Point), just north of Mutiny Bay (courtesy of the WeatherUnderground).

The bottom panel is wind direction. Wow... a VERY sharp, almost instantaneous wind shift from southerly to northerly at 3 PM.  The top panel (red line) shows temperature, which plummeted after the windshift.  Other nearby observations showed the same thing.


The visible satellite image 3 PM shows a band of clouds associated with the windshift line.


Really not that impressive and the weather radar showed no active precipitation with the windshift line (see below).

3 PM weather radar

The vertical structure of the leading edge of the wind shift line (illustrated by the figure above) would be a narrow wedge of northerly winds at low levels, with the southerly winds right above.   High-resolution UW WRF model simulations indicated such a structure (see simulated sounding with height at 2 PM for the point in question).  The Camano Island radar noted the same thing.
With all this meteorology in mind, let's get back to the float plane accident.  According to flightradar24 (below), the plane was flying around 600 ft AGL, climbed to around 1000 ft,  and then rapidly lost altitude at 3:09 PM (2209 UTC).  Aircraft speed was around 125 knots earlier and then lost speed just after 3:08 PM.


Flying south they had the northerly winds behind them.  Then they hit the convergence zone line.  Moving from northerly to southerly winds could explain their rapid loss of speed.  Although their ground speed would decline, their airspeed could increase, causing more lift.  Did that contribute to their increasing elevation at the end?

Not only did the wind reverse suddenly, but there would have been considerable turbulence due to the strong vertical wind shear associated with the convergence/windshift line.

Did this have anything to do with this accident?  

I do not know.  All I can say is that the accident occurred in an area of an active wind shift and potential low-level turbulence.  Is it a coincidence that the accident occurred simultaneously with their traversal of the big wind shift?

10 comments:

  1. This may be the first hard evidence for this crash. I expect any number of FAA and NTSB people will be looking at this. They may even be giving you a phone call. Severe turbulence hitting a decades old plane may also be of concern.

    ReplyDelete
  2. Something else to consider: odds are they wouldn't have traversed squarely into that windshift line, which -- if they could have been aware of it -- would have been the safest angle of attack (like aiming a boat into a wave).
    That surge of southerly wind would have generated sudden lift on the windward wing, while also likely yawing the plane into the wind. That's a weird, disorienting maneuver that can easily transition into at least a temporary out-of-control spin. Pilots are trained to recover from spins, but in this case they had only 1500' of altitude to work with... they would have had very little time to recover.

    ReplyDelete
  3. Is there a place the public can access weather radar archives, like the "3PM weather radar" pic in your post?

    ReplyDelete
    Replies
    1. Lefty, UW has a nice display of radar data for our area. You can choose the end time, the number of hours of radar images to display, the frame rate, and you can pan and zoom around the map to see areas of interest relatively closely:

      https://atmos.uw.edu/current-weather/northwest-radar/

      Delete
  4. We can assume the plane lost control, based on the fact that it can land on water unpowered, their was sufficient altitude to do a normal descent, and that it apparently broke up on impact. There are several incidents in the record for this plane for loss of control, from overloading and misloading to mechanical failure leading to loss of pitch control.

    These planes are flown visually (not on instruments). With a sudden updraft and high humidity it seems possible the pilot could have encountered IFR conditions and lost spatial awareness while trying to stabilize the aircraft from the effects of turbulence. His workload could have also increased by the need to decrease speed to get down to maneuvering speed. Another possibility is mechanical failure due to being overspeed for turbulent air.

    The POH for a similar aircraft states a max maneuvering speed of 126 mph which is the maximum speed at which control surfaces can be maximally actuated such as when encountering turbulence. FlightRadar24 depicts a ground speed of about 144 mph. and the weather observation at Bush Point suggests a 10 mph tailwind switching to a 10 mph headwind, but average wind speeds at 1000 ft AGL could have been even greater, and turbulent gusts potentially increase that value even further.

    If we assume a steady 10 mph tailwind transitioning to a 20 mph gusting headwind their effective airspeed could be about 174 mph which is 28% above the handbook value for flying in turbulence. This would not be ideal for the airframe.

    ReplyDelete
  5. I flew fixed wing wheel planes from Friday Harbor to Seattle Tacoma Airport for a year from May 1989 to June 1990 as I worked at Air San Juan - ChartAir, based at Friday Harbor. One day while flying a four person, high wing Cessna 172, I was heading south with a north wind behind me, then I hit a light chop in the air, and then had a slight headwind, so I had passed through the convergence zone. I had other routes, and other larger planes, but that was the only time I noticed that condition. This larger plane had higher wing loading, and was much more stable. How observant to notice this possibility. I saw a replay of an interview with a witness to the last few hundred feet of the plane travel, and he said the plane was heading straight down, turning slowly, so something must have come apart. I assume there are special cameras that can look down into the water to look for the remains

    ReplyDelete
    Replies
    1. I saw that same interview with the witness and came to the same conclusion, that something must have come apart. If this is what happened, then the question is, why did something come apart?

      Turbulence in the convergence zone might well be part of the answer to that question. The kinds of stresses that would be induced inside the airframe from a sudden encounter with strong air turbulence could be quite complex, given the presence of two large pontoons hanging from the fuselage.

      Other questions to be asked include how many hours were on the airframe; when was it originally manufactured, and how many times had it been modified given that older models of this float plane have been given a number of upgrades over the years.

      Delete
  6. As the NTSB noted and in looking at the data (from flightradar24.com), this plane really appears to dropped nose first pitching up and then straight down hitting the water violently. I agree with Allan, that something must have come apart from the structure of the airplane.

    Time Alt Speed Direction Speed FPM
    2022-09-04T22:08:18Z 900 124 151 0
    2022-09-04T22:08:24Z 900 122 150 128
    2022-09-04T22:08:30Z 900 117 149 -128
    2022-09-04T22:08:36Z 900 112 149 320
    2022-09-04T22:08:42Z 1000 108 147 1,472
    2022-09-04T22:08:45Z 1000 102 145 -768
    2022-09-04T22:08:50Z 700 80 141 -7744

    The Skunk Bay Weather cameras do point directly towards the crash site which should provide more visual context with the local area weather at the time of the crash.

    ReplyDelete
    Replies
    1. A more minimal statement would be that the plane was not under proper control during that descent. Why is the question NTSB will be searching for.

      Delete
  7. As is being discussed on the WUWT version of this article, speculation is now focusing on a possible failure of the elevator trim tab servo unit and/or its surrounding structure. A failure of that system would cause the elevator to flutter thus causing the pilot to lose control of the aircraft.

    As is noted in the YouTube videos hyperlinked from several WUWT comments, four previous crashes have occurred in this same type of aircraft because of a failure of an elevator trim tab and/or its surrounding structure.

    The actual cause will not be known until the wreckage has been retrieved and a thorough investigation has been completed.

    ReplyDelete

Please make sure your comments are civil. Name calling and personal attacks are not appropriate.