The Flying Bear Goes to Beantown | Part 4, Going Missed

What? Me Worry?

We planned to depart Rockport the morning of Thursday, July 11. Rather than going straight home, we scheduled a tour at a Maine college that interested The Bear. However, a forecast for low IFR weather in that part of Maine led us to cancel it. I had no interest in flying one hour north only to be shut out of our destination by weather below minimums for the available instrument approach procedure. Besides, who wants to wander a college campus in the rain?

So we planned to go straight home to Sodus from Beverly, assured of an easy IFR flight by a reasonable weather forecast. It did not occur to me that the scenario I had considered and rejected for Maine -- being unable to land at the destination after a cross country flight of some distance due to weather -- could come to pass at home.

Farewell Beverly

Date	Aircraft	Route of Flight	Time (hrs)	Total (hrs)
11 Jul 2024	N21481	BVY (Beverly, MA) - SDC (Sodus, NY) - ROC (Rochester, NY)	4.0	2853.1

Our experience at Beverly Regional Airport was positive. ATC was great, the FBO (FlightLevel - Beverly) treated us well and charged reasonable fees, and radar services were managed by the perennially capable Boston Approach.

Moments after climbing through the ceiling over Beverly, MA.

We made an IFR departure that morning on runway 16 and climbed above the ceiling in short order. We were assigned a slightly ridiculous alteration to the route I filed: MANCH T316 LAMMS T608 ROC that had us flying past Sodus to the Rochester VOR before setting up for Sodus. But I accepted the clearance knowing that I could change it with a local Approach control like Syracuse.

As we flew westward, the layer broke up significantly and offered glimpses of civilization passing beneath.

We passed Keene, NH, the site of my partial engine failure earlier this year.

Approaching the Green Mountains of Vermont, the ceiling began to close back up. One of our last glimpses of the ground was the Connecticut River that physically coincides with the border between New Hampshire and Vermont.

We spent some time in instrument meteorological conditions over Vermont. At one point, we entered a region of sufficient turbulence that I disengaged the autopilot to hand-fly. Caught in series of up and down drafts, the autopilot pitched the Warrior aggressively to maintain altitude and the indicated airspeed trended too high in each updraft. One particularly spectacular bump surprised all three of us as we were jolted out of our seats (even with seatbelts fastened).

I logged about 0.7 hours of IMC time on the flight. Most of the trip was in visual conditions above a solid overcast and I usually do not log such conditions as instrument time.

Unforecast

Southeast of Rome, NY we were passed from Albany Approach to Syracuse Approach. We were generally in the clear above a solid overcast layer and the air was smooth. Once on with Syracuse, I decided to rectify the inefficient routing at the end of my instrument clearance and requested direct to Sodus from the Syracuse VOR. After a pause, we were instructed to fly directly from our current position to Sodus. This was even better than I requested.

Screenshot from ForeFlight while several miles southeast of Rome, NY.

Still the better part of an hour away from home, weather conditions were concerning. Although the forecast had not given me any pause during planning, reality was that the Williamson Sodus Airport was overcast at 300 feet. Since we lost our LPV approaches due to obstructions along the final approach course, the downgraded LNAV minimum descent altitude for the approach to runway 28 was 558 feet above the ground. This meant that there would be no getting into Sodus unless the ceiling lifted.

Indeed, it did improve from low IFR to IFR conditions with a 500 foot ceiling by the time we reached Syracuse. This put us right on the edge of being able to land and I opted to try.

I had been flying for nearly four hours. I was tired and, despite a great breakfast at Red Skiff in Rockport that morning (including some of Kristy's grilled muffin), I needed to eat. It was not the best situation to initiate an instrument approach into my home field with weather conditions hovering around minimums.

Unanticipated Consequences

Earlier this year, I hand flew a couple of approaches with a safety pilot (Ed) and was extremely displeased with my performance. So I flew a few more practice sessions to tighten up my proficiency, all of them by hand. This was in contrast to my usual practice routine of flying half by hand (to maintain manual proficiency) and half by autopilot (to maintain autopilot proficiency). By overcompensating for manual proficiency, I had unintentionally allowed my autopilot proficiency to degrade. When I have passengers, I usually choose to fly coupled approaches with the autopilot because it is generally more precise than I am and that is what I did in this case.

I was cleared for the RNAV-28 approach into Sodus and set the avionics appropriately. I managed the stepdown to 2,000 feet at JAKRR (named for WFC founder Jake DeGroote) by programming the autopilot for a descent to that altitude. However, due to my mental and physical state as well as my lapsed proficiency on the autopilot, I failed to activate approach mode on the autopilot. I normally do this upon hearing, "Cleared for the approach."  As a result, the autopilot did not capture the glide path on interception. Belatedly noticing this, I disconnected the autopilot and hand flew the rest of the approach. But I was hesitant to descend too aggressively to recapture the glidepath and essentially tracked downhill while remaining above it. This was not a good setup to manage an approach down to minimums.

We broke out right at the missed approach point which is much too high and close to the runway threshold to make a normal landing. I initiated the missed approach procedure and watched the hangar buildings disappear back into mist as we climbed upward into the clouds.

I knew exactly where I had erred and reasoned that we should be able to make it in provided that weather conditions did not degrade in the interim. Rochester Approach vectored us back around onto the approach and I did everything right. As we neared the field, Chief Instructor Mike called on the radio. "The visibility has gone down to a mile and a half since your last attempt. Sometimes I have good luck switching to runway 10."

"If we don't get in this time, we're going to Rochester," I responded and was answered by an affirming double click on the mic from his end.

We reached the minimum descent altitude and the airport was nowhere to be seen. We went missed again, climbed higher into the murk, and requested a landing at Rochester. We were assigned the ILS-28 approach into Rochester and I reconfigured the avionics to accommodate that. I was tired and it was a bit of a scramble, but I got there. Conditions were about the same at Rochester, but the ILS-28 approach goes down to 200 feet and we broke out well before the runway threshold to make a standard landing. 

As the Warrior's wheels touched down on the runway, I was mentally finished. I was assigned a multiple step taxi clearance to AvFlight - Rochester and had to ask for the last step a couple of times because my brain just could not hold all of the information flooding in by radio.

My friend Bob manages AvFlight - Rochester and I knew from him that they do not charge single engine GA airplanes all of the fees that their on-field competition, US Airports, does. We were marshalled to parking within view of an F-22 Raptor visiting for the National Warplane Museum's pending airshow -- the so-called Greatest Show on Turf. We were graciously received at AvFlight and Bob offered a courtesy car for overnight use.

"Hey Bob, why is that Raptor sitting on the US Airports ramp instead of yours?" I asked him teasingly.

"Well, that's because US Airports promised a hangar for the -22 and I could not guarantee the space," he explained.

"But it's sitting outside in the rain right now," I needlessly pointed out to him. (Tired, remember?)

"Yep," he acknowledged with great restraint.

While speaking with Bob, I met a young pilot who also worked at AvFlight. "You have a blog, don't you?" he asked me.

I was so tired and so surprised to be recognized that I did not even know how to react. "I do," I answered with maybe too matter of fact a tone.

The first order of business was lunch. I drove my family to Naan-Tastic, a quick casual restaurant based on the Chipotle model, but for Indian food. It is a favorite of The Bear's and an inconvenient distance from home, so we do not go often. After a long, turbulent slog from Boston that ended with three instrument approaches in actual IMC, it felt like the least I could do. With bodies refueled, we drove ourselves and our luggage home.

Collapsed in a recliner in my own living room, I immediately began planning a repositioning flight to get the Warrior back into her hangar.

Repositioning Flight

When we arrived at Rochester in the rain, it appeared that the best plan was to move the airplane the following day. However, by the time I was rested and reinvigorated after lunch, it was clear that weather conditions around the area were improving. Still IFR, but barely. The ceiling lifted to 900 feet and even began to break up over Rochester. Repositioning the Warrior that day would simplify logistics for the next day, particularly because it would make my car (presently sitting in my hangar where the Warrior should have been) available for work the next day.

Date	Aircraft	Route of Flight	Time (hrs)	Total (hrs)
11 Jul 2024	N21481	ROC (Rochester, NY) - SDC (Sodus, NY)	0.7	2853.8

I think Bob was surprised to see me reappear that afternoon. I thanked him for being such a great port in a "storm", pointed out that the check engine light was lit in his courtesy car ("Oh, it's been that way for months," he said), and attempted to pay my $3 county-mandated landing fee. Seeing me pull out a credit card because I was out of cash (unless Avflight wanted a portion of the 30 euro still in my wallet from Italy), Bob took the cash from his own wallet and paid on my behalf. Now that's a great host!

Moments after departure from runway 28 at Rochester.

I departed runway 28 from Rochester with minimal fuss and started the turn on course before climbing up through the ceiling.

Unlike the misty air prevailing at the time of our arrival in the area, the atmosphere was significantly crisper and clearer. Current conditions at Sodus were high enough for a successful instrument approach, but still low enough that the approach could be logged. I set up for my third attempt at the Williamson Sodus RNAV-28 approach.

Unlike Rochester, it was still gloomy over the field, but the ceiling was obviously much higher as I flew past Sodus Bay. I landed without issue, though the wind was squirrely and the touchdown was not one of my best. Still, it was good to be completely back home. I took on fuel from the temporary fuel system put into place during our fuel farm rehab and went in search of Mike.

"I saw you pass over the field both times," he affirmed about my earlier approach attempts. So close and yet so far.

Mixed Bag Reflection

ForeFlight ground track showing two approach attempts into Sodus and the successful one into Rochester.

I learn things on almost every flight, but some experiences inspire more learning than others. In this case, I think my performance was definitely a mixed bag. 

While I have flown four hour legs with successful outcomes before, I had never shot an approach to minimums after such a long leg. I need to do better about having food on board so that I can be at my best when it's time to dive into the murk so close to the ground. That is not always easy or convenient to arrange, but there was nothing preventing us from stopping at a store in Massachusetts before taking off that morning.

I also reinforced something that I already knew intellectually, but now understand viscerally: autopilot proficiency is just as important as hand flying proficiency when it comes to conducting instrument approaches. The GFC-500 autopilot is fantastic, both capable and intuitive and I am still very pleased with my decision to install it. But like any programmable device, it does not read minds and only does what the operator tells it to do. In this case, I did not activate approach mode to capture the glide path in a timely manner. That is completely on me. I lacked autopilot proficiency because I overcompensated for some poorly executed hand flown practice approaches and eschewed practicing with the autopilot in favor of doing everything by hand. I should have included an ILS or two with the autopilot in my last practice session.

I am convinced that lost autopilot proficiency exacerbated by being tired and hungry directly necessitated the diversion to Rochester. Without that combination of elements, I think we would have landed safely in Sodus on the first attempt.

The value of attempting the same approach a second time is one of those thought experiments that aviation safety experts love to debate. To my mind, the likelihood of such a thing working out is relatively low unless the weather is actively improving. In this case, I think that my second attempt was warranted because I had diagnosed what went wrong the first time around and understood that, if flown on glide path (instead of above it), I would have been in a position to land visually before reaching the missed approach point. The worsening of the weather made the difference in this case, but I still think the thinking was sound. I had plenty of fuel, too. With full tanks and a leaned fuel burn of roughly 8 gal/hour in cruise, the airplane clearly has longer legs than I do.

In the eleven years elapsed since earning my instrument rating, this was the first time that I was unable to complete an approach due to weather being below minimums. In this circumstance, some pilots are tempted to cheat and descend below minimums if they think (correctly or otherwise) that they are close to breaking out. I did not do this and am pleased that my immediate instinct was to fly the procedures as written rather than succumbing to a temptation to cheat. This is especially true because I had my entire family on board. Sometimes we do not know how we will behave in a stressful moment, but I think this was a good example of training having established the right mindset.

In the final analysis, it feels like I did not set myself up well for success: a long flight, inadequate food, and a little rust on flying approaches with the autopilot. However, when that first approach attempt fell apart, I managed it in an appropriate, by the book, manner.