Low Visibility Takeoffs: How Low Is Too Low?

I spent most of my career flying out of Westchester County Airport (HPN) in New York. The airport is perched 439 feet above Long Island Sound, so fog and low visibility are common. HPN is always the first airport in the New York area to crump when low weather moves in.

When the runway visual range (RVR) is low, HPN ground control starts asking departing pilots what their takeoff visibility minimums are. The reason is so they can stack the taxiway with the crews having lowest takeoff minimums first. So the pilot with the 600 foot RVR requirement goes to the head of the line and is cleared to takeoff as soon as the RVR reports that value while pilots with higher visibility requirements wait in line for improvement.

What Are Takeoff Visibility Minimums?

Takeoff minimums depend on the rules you are flying under. Operating certificates required for airlines and others who fly for hire will specify takeoff minimums, often for each airport or runway. Other certificate holders may have general takeoff minimums the apply broadly. If you fly for your own reasons under FAR Part 91 takeoff minimums are up to you, though most corporate flight departments set departure minimums as part of their standard operating procedures (SOP).

Foggy runway
How low is too low?

But, you say, charts sometimes show takeoff visibility minimums. True, but FAR 91.175 that addresses takeoff minimums only applies to operations for hire. The rule is silent on what is sufficient visibility for a not-for-hire departure. It’s up to us who fly for personal and business reasons.

The lowest routinely used takeoff visibility minimum is 600 feet RVR, which is about one-tenth of a mile. Pretty low. It’s the lowest visibility typically used in simulator training for taxi and takeoff practice.

If no takeoff minimum is listed for an airport or runway then “standard” minimums apply to those flying for hire. For single and twin engine airplanes the standard departure visibility minimum is one mile.

Many pilots flying under Part 91 believe using the approach and landing minimum visibility as a takeoff minimum makes sense. The logic is that if you need to return immediately, for whatever reason, you can expect to successfully make the instrument approach back into the departure airport.

Many general aviation airports lack weather reporting so it’s impossible to precisely know the prevailing visibility, much less RVR. Experienced pilots I know have developed their own minimums such as being able to see the end of the runway. Or see some object at a known distance from the runway. Or even some want to be able to see a set number of runway edge lights before they go.

What Are The Risks?

The big risk that jumps to mind is engine failure during a low visibility takeoff. And that would be a critical situation. But the accident record shows that is an extremely rare event. Given that engine failure itself is uncommon, and that low viz takeoffs are infrequent, the odds of an engine failure during the seconds or couple minutes of a low viz takeoff are very long.

The very real risk—confirmed by the accident record—in making a low viz takeoff is loss of control. Obviously, you need to see far enough ahead to steer the airplane down the runway while accelerating to rotation airspeed. Departing the edge of the runway will be bad, but more likely survivable than loss of control in flight after liftoff.

Transitioning instantly from looking out the windshield to maintain centerline during the takeoff roll to flying only on instruments as the airplane rotates is one of the most demanding tasks in IFR flying. And our natural balance and sensing system makes it much more difficult.

The somatogravic illusion created by our inner ear balance system makes acceleration feel like a climb. Lifting off into the fog will bring on this powerful sensation, making any pilot believe the nose is pointing higher and higher. The pilot who can’t suppress this illusion and believe only the flight instruments will sink back into the ground believing for all the world that the airplane is pitching up into a stall.

Our sensory system is constantly feeding us misleading information while flying on instruments, but it’s the instant transition from the visual world necessary to maintain centerline on the takeoff roll to solid instrument flying that is the big challenge.

How To Minimize Risk

The most critical risk mitigator for a low visibility takeoff is to be a competent and confident instrument pilot. The confident component is really vital. Making a low visibility departure is no time to be wondering if your skill is up to the task.

Another critical risk element is the takeoff runway. How wide is it? How bright are the paint markings? Are the lights bright and closely spaced? If you can’t see to stay on the runway until rotation the best instrument flying skills won’t matter.

Instrument panel in Cessna
The instruments are always there, but you have to trust them.

Is the airport controlled, or at least secured? Many GA airports have little control to vehicle access so can’t you be sure some pilot isn’t driving to his hangar down that foggy runway. Or that another pilot hasn’t announced his intentions to enter or cross an uncontrolled runway. And then there is always the chance wildlife is hidden in the murk ahead. Are the fences at the airport high and secure? An operating control tower can’t prevent all runway incursions on a foggy day, but it sure is a huge help.

Do you have a flight director? In larger airplanes no pilot takes off under any weather conditions without the flight director command set to the initial rotation and climb pitch target. Under the high workload of takeoff, the flight director command bars give you a single point to check in your scan. If the pitch attitude is in the bars the airspeed will be within target range and you will be climbing, no matter the lies your sensor system is telling you. I just wouldn’t make a low visibility takeoff without a functioning and capable flight director.

What About A Single Engine?

When you fly single engine airplanes IFR or at night you have already accepted a degree of risk to reap the reward of much lower costs and availability a single provides. Does making a low visibility change that risk much? I don’t think so.

Unless you fly IFR only with good daylight VFR below you, loss of power in a single while in the clouds will require some luck to survive the forced landing. Your opportunity to try to maneuver at the last minute to find a more hospitable landing spot is quite limited even with a mile or two of visibility.

If you make a low visibility takeoff, other than the immediate risks of loss of control on the roll, or shortly after liftoff, I don’t see any added risks compared to making an IFR departure with standard takeoff minimums. You’ll be in or on top of the clouds after the low viz departure, which is no different than if you departed with a mile visibility. En route the viz and ceiling will undoubtedly be different than the departure airport, so that’s another discussion independent of the low viz takeoff risk.

Having an alternate airport with weather above approach minimums within reasonable range makes complete sense for multi-engine airplanes. For the single-engine pilot the alternate is important, because you could have conditions other than power loss that require diversion. But your alternate planning for a low viz departure is really no different than what you’re comfortable with for any IFR flight.

As I see it, the low viz takeoff is no more risky than any other flight in instrument conditions if you are confident you can see enough to stay on the runway until liftoff, that the airport environment is reasonably secure so there’s no obstruction on the runway, and that you can instantly transition to very hard instrument flying.

Those are all big ifs. And the rules for those flying for their own reasons allow us to come up with our own answers to the risk questions.

22 Comments

  • Happy to see a reasonable answer to this question. I am tired of people saying low vis takeoffs are dangerous.

    My rules are, can I see a few runway lights to maintain direction until liftoff?

    I have taken off a 150’ wide runway where the runway lights were too far away laterally. I side stepped and put center line lighting out my left side window. Accelerated and rotated. Did not even seem very difficult.

  • Military has a cliche about giving “special” cards for 0-0 takeoffs only to those smart enough to not use them…aka, just because you can, doesn’t mean you should.

    Risk has two components, likelihood and consequence. Airplane doesn’t know when it’s dark, low viz, over water, but each of those make for less pleasant consequence for loss of your only engine. I have also seen highly capable redundant avionics installs done by “Joe” in the local avionics shop taken out completely by a single bad ground.

    My advice when this conversation comes up is to advise circling mins…and ask what lifesaving mission the flight is supporting.

    The outcome for a highly trained crew in a well engineered, multiple-redundant aircraft is probably going to be different than for the typical Part 91 bug-smasher/single pilot… as PIC make damn sure any passengers you put in that situation understand that difference.

  • During my Part 135 check hauling days, we were certified to use the lowest available Cat I straight-in approach minimums (down to 1800 RVR). The transition from following the centerline to instruments and dealing with vertigo brought on by the acceleration was the biggest challenge.

    We flew in all weather, five days a week with 6-month check-rides. All our pilots stayed on the top of their game. The skills needed to safely accomplish these kinds of operations are perishable. Low visibility/ceiling risks can be well managed when the pilot is committed to keeping up such skills. But if not, it is best to stay home.

  • What about synthetic vision? On a recent IPC, I did a simulated 0-0 take off under foggles at the beginning of the take off roll. No transition needed, you’ re already on instruments. Seemed to work well.

    • Hi Max,
      During my most recent recurrent simulator training at FlightSafety/Textron I wanted to fly the standard departure with an engine failure at V1 out of Bozeman (BZN), an airport I fly into and out of several times each winter. We’re always at max takeoff weight, and the performance info shows the King Air 350i is challenged.
      The airport detail wasn’t in the simulator visual data base, but the procedure was, and the terrain was. The King Air successfully met the climb criteria engine-out and I returned for landing at BZN. There was no visual representation of the runway through the windshield, but the synthetic vision images on the PFD were perfect and an exact image of the BZN runway.
      By looking only at the syn viz, and flying the ILS, I made my best landing of the entire recurrent sim course. So, yes, I think syn viz can be a great help in terrain avoidance and even low viz landing when nothing else is available.
      Mac Mc

  • Questions I would have…is it certified in that install for that use? Have you/your CFI trained to all the failure modes/mitigations and understand the outcomes, can you and are you willing to stay proficient?

    Is the synthetic vision a factory install with many engineering/test hours of both the design, equipment and installation…or a local install to a 50 year old wiring harness/power distribution system? One of those may have a 99.9999 reliability, the other? For the local install, did the processor react well to the alternator transient as the engine went to max rpm, did it tolerate the expansion joint jostling down the runway?…a momentary hang at altitude is no big deal to clear, on the roll…

    While system failures can occur anywhere, failures close to the ground give few options or opportunities to mitigate. Tesla “autopilot” crashes are a good example of “off design” employment leading to bad outcomes.

  • Ok, my organization requires 200 foot ceiling and 1/2 mile visibility for departure unless the PIC has 50 hours of actual weather time, in which case zero zero is authorized. My thought is if you have 50 hours of weather time you should have more sense than to takeoff zero zero. I use the minimas for the approach.

  • Thank you for your well reasoned and realistic comments on low visibility operations in single engine airplanes.

    Takeoff speed, climb angle and somatogravic effects are modest in the low performance world (P-factor can still be an issue) and, as you indicate, low visibility takeoffs can be viewed as part of the normal repertoir for a capable, current single engine pilot.

  • Congrats Mac, you’ve pulled the pin and rolled another one into the room.

    So all the folks who’ve had the experience, equipment and training to do this in the past are saying “no thanks, no good reason to do this and no longer have the recurrent training or right equipment”

    The folks who think this is a good idea missed all your caveats and only saw “Mac says this is no worse than any other IFR flight”.

    I have no issue with the sole occupant exercising their Part 91 privilege to do this, what I do have a problem with is that if they take pax along, they probably won’t brief “if my engine, avionics or training fails, you will all soon be dead”. Those same pax would probably never hop aboard for a ride out of a one way strip facing a rock face or low altitude aerobatics…but what’s a little fog?

  • Interesting! In 2001, Feb. to be exact, I was working with Peterson Flying Club, Colorado
    Springs, CO. After three weeks, I wanted to return home to KS (K78). Upon rising the ceiling was 0 and visibility about 200 feet. At 8 AM, “it will come up in a couple of hours”. At 9, then 10, then 11, no change. I returned to the Guest housing for another night.
    Saturday 7 AM, Wx was about the same. By 9, I called home and found that the Wx there
    was 4500 scattered and better to the East. I called FSS and filed an IFR departure with the lady taking my flight plan not even cautioning me. Called the tower, got the clearance and proceeded to pick my way to the runway. On the center line I announced ready for departure and cleared for that. Upon rolling, I barely had one center strip in
    my view at a time. My Mooney accelerated, got light and began to climb. cleaned up
    the airplane and less than a minute later I was above the fog.

    I really did not think much of it until I switched from Denver to Kansas City Center and
    the new operator asked if I had heard from a specific N number, a 190 that was overdue into the panhandle of OKlahoma? I had not but gave a few calls in the blind
    just in case I could help. This is when the big question hit me, “what if I had an engine
    out on take-off?” Kind of humbling, but it was nice to sleep in my own bed that night.

    Would I recommend that to anyone, probably not, but then I would have to ask about the experience level, and reason for wanting to. I have done this with several of my students, though under the hood. Just do not try something like this if you are not comfortable with it , not familiar with your equipment and is it really, really necessary.

    I have also practiced zero-zero landings (with hood) because that too might just become a requirement though not planned. Stay healthy and thanks for the article.

  • I recently did a few 600RVR takeoffs at recurrent (Falcon 900LX) and, as long as nothing goes wrong, it’s a piece of cake. The left seat is outside the windshield while the right seat is monitoring the synthetic vision and LOC needle, if available. Things get really busy really quickly if something goes wrong, though. If an airplane pulls out on the runway in front of you just before v1, there’s almost no time to react. V1 cuts can be interesting because if the centerline isn’t maintained when the engine fails, it can be difficult to find the middle of the runway again, not to mention having to shoot a One-Engine Inop approach while hoping to get a visual on the rabbit, then the runway to make a safe landing.

    All that said, our (Part 91) Flight Operations Manual says we can takeoff in 600RVR as long as we’re trained, but we all agree that we won’t go unless the weather matches the minima on the ILS or LPV approach.

    As with many things in aviation, low-vis takeoffs aren’t a problem until they are. Great article, Mac!

  • In my area the lowest visibility takeoff days are usually the nicest IFR flying days as once through 1000′ you are in the sun, while the days when visibility isn’t an issue involve pounding along in cloud for the entire flight.

  • We trained for this at “Sunken Lunken” (KLUK) as fog was extremely common, lying in a valley large rivers on two sides. Not that I would do it, but because it was technically legal, the instructor wanted to make sure that you could maintain directional control, and climb out and turn as necessary.
    This was always done with zero outside visibility, at least for me: under the hood! A bit of fun, but knowing that the instructor would keep you out of the lights helped.

  • Great article, Mac. HPN became my home airport earlier this year and I have been lucky enough (as of now) to not have dealt with a fog or low takeoff minimums and the departure delays that result. Thanks for your insight on this topic. I find that most pilots on this subject quickly go to “don’t ever do it” but I believe there’s more nuance than that.

    I don’t know if you ever still come through HPN but hoping for the chance to meet you someday. I’ve been reading your stuff since the 90s when I was a teen and first developing my love for aviation!

    • Hi Doug,
      Thanks for the kind words. I haven’t been into HPN for several years. But one thing that I’m sure hasn’t changed is the instrument departure from Runway 16.
      Starting a 160 degree turn at 400 feet agl then leveling at 3,000 after a low-weather liftoff is one of the more demanding bits of instrument flying I can think of.
      HPN is one of the great bizav airports in the country.
      Bests,
      Mac Mc

Leave a Reply

Your email address will not be published. Required fields are marked *