Monday, October 17, 2011

Into the Darkness

Hindsight being consistently 20-20, I now find myself looking back on last Friday's flight from Bridgeport (KBDR) to Potomac Airfield (KVKX) with acute interest. The interesting bits came in three distinct parcels; I'll describe them in turn.

Part I

The previous post reviewed my planning for the flight, and included an off-hand comment that it's the time of year when one needs to think "less about, 'Will there be convection issues?' and more about 'Will there be ice?'" Famous last words.

As N631S and I were taxiing for departure, the controller advised that ATC was not providing IFR releases for aircraft headed west or south "due to weather." I said I'd park and wait it out. On the way back to the tie-down I took the screen shot at left, showing the weather that was causing the problem.

I was aware that there was significant weather in that area, but I had felt I could depart and work with ATC to find a route through the line or at worst, turn around and return to KBDR. But that scheme didn't work for ATC! In the busy New York Approach airspace, I guess that there isn't time to work individual aircraft through a line of convection. They just shut the routes down until the weather improves.

So, N631S and I sat on the ramp, checking in with the tower every 15 minutes for an update. After an hour and a half, the answer came back as, "Things are improved over by Sparta...you'd better get taxiing." Which we did!

The screen shot at left was taken soon after departure and shows the break in the line that we were aiming for. With just a few vectors for spacing, we were on our way westward. There were some cloud buildups to go through that looked a little intimidating, but the NEXRAD radar display showed little precipitation and there was no indication of lightning. And in the event, the clouds contained nothing worse than light turbulence.
At left, the view at 8,000 feet, coming out of the far side of that line of weather near the Sparta VOR (SAX). You can see the glow of sunset, a result of the delay in departure from KBDR. I hadn't expected to be logging night time on this trip but it was working out that way.

Part II

In looking at the weather forecast for this flight, it was clear that there might be some flirtation with ice. And near Allentown at 8,000 feet N631S and I found ourselves in the cloud layer with the outside air temperature (OAT) coming down. I wanted to stay at 8,000 as long as possible because the winds were more favorable there than at 6,000. Despite a true airspeed of about 140 knots, speed over the ground was only about 110 knots. Lower, the wind would be more directly "on the nose" and the headwind component stronger. I watched the declining OAT until it reached 34°F, and then asked Allentown Approach for a descent to 6,000. And as expected, about 7 knots of ground speed went away. But that altitude was below the clouds and significantly warmer.

Except for needing to be vectored around an isolated patch of convective weather just north of Lancaster, the balance of the en route portion of the flight was uneventful.

Part III

As had been forecast, the surface winds in the DC area were strong and gusty. And, as expected, the winds at KVKX were moderated by the field's location in a valley. Still, there was enough wind (reported as 7 knots from 260°) to make Runway 24 the clear choice. My final controller from Potomac Approach asked if I had the weather at KVKX and I replied, "Yes, I've picked that up; looks like it'll be Runway 24 tonight."

That elicited an offer that I couldn't refuse. Andrews AFB, which lies about 5 miles northeast of KVKX, was more or less between my position and the approach end of runway 24. I could get a turn toward the airport passing just south of Andrews, provided that I agreed to timely cancellation of IFR (since I'd wind up below Approach's Minimum Vectoring Altitude). Visibility was good, so I said, "Sure, we can do that." To which the controller said, "Proceed direct to VKX while I talk to Andrews."

A few minutes later I was at 1,500 feet, essentially crossing east to west over the approach lights of Andrews' Runways 1R and 1L. I was set up on a left base for 24 at VKX. Having cancelled IFR when I had the beacon in sight, I reported the runway in view and was released by the controller. N631S rolled out on about a 3 mile final.

It was a bit after 8 PM local time, and completely dark. Approaching KVKX from the northeast – something I'd never before done at night – offers only the most sparse ground lighting. The runway lights are clearly visible and welcoming, but the intervening terrain is something of a "black hole". The area is basically flat and featureless and I was down at about 1,300 feet to stay well below the floor of the Class B Airspace.

We've all read the training materials on visual illusions. We've been told about the dangers of the "black hole approach." (Avoiding Black Holes by Dale Wilson gives a good overview.) But now I was about to have the experience.

The visual approach aid for Runway 24 is a two-light VASI (Visual Approach Slope Indicator). Red beside red means you're low. White beside white, you're high. Red with white is just right. As I looked out into the dark night, the VASI was stubbornly red with red. Yet everything else about the sight picture had my ground-dwelling brain screaming at me, "Too high! You're too high! Get down!" I had to consciously remind myself, "It's an illusion. Do not descend. Trust the VASI." I was applying that first rule of IFR flight – trust your instruments! And soon, one of the red lights turned white and I started a descent toward the runway.

The landing itself was uneventful, and I was left reflecting on how a VFR-only pilot, trained to trust his eyes, could so easily be trapped by what I'd just experienced. It's a sobering thought.


Two nights later. Sunday, about 8 PM, N438CP – a Cirris SR-22 – was on a visual approach to Runway 26 at Danbury (CT) Municipal Airport (KDXR). The weather was good. I've flown into Danbury on occasion and can state that the terrain surrounding the airport is interesting at the best times.

The approach to 26 is over a residential neighborhood, presumably fairly dark. About a half mile short of the runway, at a point approximately 160 feet above the threshold elevation, the Cirrus flew into the terrain (reportedly where the red dot appears at left). The pilot, sole occupant of the aircraft, was fatally injured. And now I wonder...was he a victim of the siren call to fly lower that I was fortunate enough to be able to ignore?


Anonymous said...

Big fan of your blog! As I pilot out of KBDR I really enjoy reading about you IFR planning and actual flights. DXR terrian is little unique for the Northeast and a sad day when any pilot FITs. Night IFR into a possible unfamiliar airport with terrain is beyond my comfort zone.

Keep up the good work.

Frank Van Haste said...

Dear Anon.:

As you say, night IFR into an unfamiliar field can be a little edgy -- but it's a lot better than night VFR into the same field!

Thanks for reading and commenting. Keep the shiny side up!


Royski said...

Thanks for the writeup. According to the AF/D, the VASI for runway 24 is set up with a steep 5.5 degree approach, which probably contributed to the illusion.

Frank Van Haste said...

Hi, Royski!

Thanks for commenting. I agree that the steeper-than-"normal" glide slope could exacerbate the illusion. Another intensifying factor is the short runway.

Best regards,


Chris said...

Frank - one thing that makes that FIT accident so hard to understand is that it was a Cirrus. I would assume the Avidyne had terrain awareness. I understand there are other factors, but it should have been an item in the toolbox specifically for such a situation.

Regardless, I'm a local pilot who flies out of HEF or JYO, because I've always been a little wary of KVKX. Don't want to take away the focus of this post, but I'd be curious to hear your take of a "KVKX for the low time pilot" theme sometime.

Frank Van Haste said...

Hi, Chris:

Do you remember that CFIT out west a couple of years ago that involved a CAP 182 and two very senior pilots? That a/c was equipped with a terrain-aware G-1000 system and on a perfectly clear night they flew it into the side of a mountain.

In the Danbury case, visibility was good and the terrain was unthreatening. It's quite possible that the terrain feature was not selected, and it could have been suppressed in landing mode.

With regard to VKX and the less experienced pilot...well, there is an active flight school there. The students seem to do OK. Since you raise the subject, I'll have to think about a blog post on it.

Best regards,


LarryPetro said...

Hi, Frank. You say that you kept your altitude at 1300 ft. to stay under the Class B during the approach from the northeast to RWY 24. Were you not in the surface portion of the Class B as you crossed over the approach end of ADW RWY 01R/L? Close in to RWY 24 approach you've passed into the 1500-ft floor of the Class B. So, my question is whether you really needed to be at 1300.