4 ms·
Some thoughts from a commercial, instrument rated pilot who works in tech now but is still a current pilots and has flow aircraft similar to the Beech 350 comme
by throwaway_SiX2N 5y ago
Some thoughts from a commercial, instrument rated pilot who works in tech now but is still a current pilots and has flow aircraft similar to the Beech 350 commercially in the past, with lots of flying in the US West into many similar airports to Hayden (although I haven’t ever flown into Hayden). This is mostly facts to help spread aviation knowledge, with just a few speculations at the end.
While it’s true that the Hayden airport is uncontrolled, it is in an area in which there is class “E” airspace all the way down to the surface. This means that aircraft must be able to fly at least 500 feet below the lowest cloud to use the airport under “figure it out yourself” visual flight rules (VFR). Just one minute before the accident, the airport reported an overcast ceiling at 500 feet above the ground, so there was not sufficient weather to operate VFR, and therefore all flights would have been Instrument flight rules (IFR).
At an uncontrolled airfield in IFR conditions, it’s true that aircraft have to coordinate with one another for ground operations (who’s taking which taxiway, etc), but must have an IFR clearance from ATC to approach and depart. Since the class E airspace goes all the way to the ground around Hayden, this means that the airport is effectively controlled, except for driving around the surface.
What this means is that every arriving aircraft needs an ATC clearance before flying the approach and must inform ATC that they’ve arrived before ATC can clear another aircraft. Similarly, every departing aircraft needs a clearance before they leave and ATC should not issue a conflicting clearance until after the aircraft is inflight and has checked in. Of course, a takeoff may not be possible, so a departure clearance is issued with a “valid time” and a “void if not off by” time. What this means is that you will receive a clearance which has a start and end time, if you cannot take off within that time window then you need to call ATC back and get another clearance.
At many remote airports, ATC would be contacted literally via cell phone (by pay phone if that doesn’t work – better run back to the plane through the snow and get your engines started real fast before your clearance expires if you only have the pay phone option – I’ve been there). At Hayden, however, there is a remote transmitter (RCAG). This means that Denver Center, the overlying air traffic control facility that controls a huge swath of the Rockies, has a transmitter located nearby so pilots can talk directly to them on the radio. As a result, taking off from an airport like Hayden in IFR is very similar to taking off from a controlled airport, the only real difference is that the actual movement around the taxiways is at the pilots own recognizance.
At airports like this without a local tower, ATC generally has to operate with a “one in, one out” procedure, meaning, essentially, that only one airplane can use the airspace at a time under IFR flight rules. If you’re on the approach, no one else can fly an approach or departure, and vice versa. You typically talk to ATC until they “cut you loose” to talk on the local advisory frequency.
In terms of runways, Hayden is in high, mountainous terrain overall, and although departure procedures are published to avoid terrain, they are very simple, and indeed, I will summarize them as climb to a specified altitude and then off you go. In addition, there is no specified runway to use in calm conditions and the weather report at the time of the incident reported no wind. This means there is no particular reason to prefer one runway over another, and in fact it would have be perfectly reasonable to select either runway on this day.
One interesting feature of Hayden is that the only approach to Runway 28 that was suitable for the weather (RNAV (GPS) 28) has a nearly 20 degree offset from the runway. This means that as the Beech was approaching the runway, it would have been coming from the South and making a final course correction just moment before touchdown, after breaking out of the clouds at 500 feet. In fact, the minimum altitude on this approach is a fairly high ~400 feet above the ground. This means that except for a short window while the Beech made it’s last 500 feet of descent (and a low 20 degree turn to line up with the runway), it would have been been in the clouds, and thus it is unlikely either aircraft could possibly have been in visual contact until very close together.
In terms of the weather, the weather report indicates it was cold, with a 500 foot ceiling. This was indicated as an overcast ceiling, meaning it fully covered the sky. In addition, the weather reported that snow ceased to fall just 5 minutes before the incident. Thus, the cloud cover has to be fairly dense to have just generate falling snow. As a result, these clouds would likely have fully obscured the aircraft from one another.
Hayden has two interesting notes in the airport directory: first, it indicates that all aircraft should announce their departure or arrival intentions on the advisory frequency 10 minutes before takeoff or landing. It also notes that departing on Runway 28 (that is to say, departing in the same direction as the Beech was flying) requires aircraft to make a turnout as “soon as safety permits” to avoid noise over the town of Hayden.
Here cease facts and begin commentary.
Based on weather, their flight patterns and operating rules for the JetBlue flight as well as where the Beech had been flying, both aircraft should have been (and presumably were) on IFR flight plans. As the Beech was arriving from the East and the weather allowed it, it was perfectly reasonable to request the approach to runway 28 (facing to the West) in order save time circling around. Similarly, the JetBlue flight wanted to head East, so a takeoff on runway 10 was reasonable. Furthermore, I would tend to see a note requesting a turn “as soon as safety permits” as shorthand for “there’s a nosy neighbor who will call the FAA and complain about your tail unless you make a very low altitude turnout”, something I would prefer not to do in the clouds while going through all the rest of the rigamarole of an IFR departure in the snow. No thanks, I’ll take 10, please.
My personal expectation flying into this type of airport is that the controller would not cut me loose to talk to the advisory frequency until just a few minutes before landing. Of course both of these airplanes had multiple radios and it should have been possible to listen and talk on the advisory frequency while still talking to center. Nonetheless, especially in certain IFR conditions like this where everyone should be on an IFR clearance, communications on the advisory frequency can often be deprioritized to focus on communications with ATC.
In any case, while talking on the advisory frequency should have been a way for these aircraft to become aware of one another on their own, they were both operating in an IFR environment, meaning that only one of them should have been in a position to be flying through that airspace at once.
This is a very scary event because it was a very near miss that should not have happened and suggests a real breakdown in the system of IFR flying at that time and place.
One small speculation: I wonder if the fact that the Beech was on the approach with a 20 degree course offset saved things? Although 20 degrees is not a huge angle, it may have been just enough to prevent disaster.
- FabHK 5y agoYes. One take-away: untowered class E airports in IMC (instrument meteorological conditions) are controlled, but the controller is remote and doesn't necessarily have radar or sees what's going on, but allocates time slices. Basically TDMA (time division multiple access) I assume the jet was flying IFR, the question is whether the Beech was on IFR and whether/when they cancelled. Let's see what the report says.