Low Altitude Alerts

I was about two miles from the threshold of Runway 27 at Oshkosh. It was a great morning to fly with high thin clouds, good visibility and almost no wind.

I had been cleared to land several miles from the airport because there just wasn’t much, if any, traffic in the area. Then the controller called and said “check your altitude, altimeter setting is 29.81.” In the background I could hear a klaxon sort of warning tone sounding in the control tower.

I asked the controller if he had received a low altitude alert on me and he said yes.

The electronic glideslope on the LNAV with vertical guidance GPS/RNAV approach to Runway 27 was exactly on center. I’m not a perfect pilot, but the flight director was coupled to the approach so you don’t need to be perfect to fly the little airplane into the V-bars and be exactly on course.

There is a four-light VASI on Runway 27 and it clearly showed two whites and two reds indicating I was on the centerline of the visual approach slope guidance, too.

This was at least the fourth or fifth time I had received a low altitude alert from Oshkosh tower controllers while flying exactly on the glideslope centerline to Runway 27. Every time I have told the controllers I was on the glideslope and had been all the way down. This time the controller said that the warning was a “little sensitive out there.” I guess so.

On a perfectly clear morning the low altitude warning was not even a distraction for me, and I doubt for the controller either, except he had to call and tell me about it. But what if I had been in the clouds? Would I believe my cockpit guidance and continue down the glideslope? Or would I believe the low altitude alert system? Now that would be a real distraction at a time when you really don’t need it on an instrument approach.

The FAA’s low altitude alerting system has been around for many years, and I believe is operational at all approach control radar facilities. I’m sure the system has prevented many accidents and is a wonderful backup for we pilots who could lose focus and wonder off below a safe altitude, and also as a backup for potential mistakes in setting the altimeter or managing our avionics.

But the algorithms necessary to calculate the low altitude alerts are complex. If an airplane never had to get close to the ground to land, the task of warning of low altitude would be simple. Just monitor the Mode C altitude of an airplane, compare that to height above the terrain below, and warn if the distance to the ground is less than 1,000 feet, or some other safe margin.

Things get complicated for altitude warning on approach because the airplane must descend, and eventually have no altitude above the ground at all in order to land. To provide useful alerts the system must determine if the airplane is flying a safe approach path, but also must consider the rate of change in altitude above terrain. If the airplane is descending too quickly the alert must be issued before the airplane is too low or there won’t be time for recovery. But the system can’t know if the pilot will arrest the descent at a safe altitude, so it has to err on the safe side and issue a warning.

These same calculations are performed by the ground proximity warning systems (GPWS) required in all larger airplanes, and the terrain awareness and warning system (TAWS) included in many GPS navigation systems for GA airplanes. But the onboard equipment has an advantage over the controller’s radar because the GPWS and TAWS can calculate the low altitude threat at least once a second, more likely several times a second. The controller’s radar alerting system gets a new position and Mode C altitude report only every few seconds as the radar scans. And the radar system only gets altitude reports rounded off to 100 feet, and our Mode C accuracy tolerance is actually plus or minus 300 feet if all allowable errors fall on the same side.

Compare that to the onboard systems that are getting WAAS aided GPS information, and radio altimeter information in the case of GPWS, continuously. And that altitude data is accurate to within a few feet, not hundreds of feet. That’s why the onboard systems can generate the warning “sink rate, sink rate” if your descent rate is higher than the normal approach path. The controller’s system is also looking at sink rate, but is getting comparatively little information so it can’t be as precise in generating a warning.

Most of the time the low altitude warning system is only active for pilots flying in the IFR system or receiving advisories with a discrete code set in the transponder. When pilots are flying and squawking VFR the system usually assumes the pilot sees terrain and has reasons to be flying at whatever altitude he chooses so warnings are not typically issued.

Even though I have excellent TAWS warning from the Garmin GNS530W in my panel, I’m glad to have the controllers watching out for me, too. But I hope Milwaukee approach, which is the facility that covers Oshkosh, can get the low altitude warning adjusted for Runway 27. I don’t mind the false alerts one bit on a nice day, but I sure don’t want to deal with them in the clouds that will be rolling in often now that winter is approaching.

This entry was posted in Mac Clellan's Left Seat Blog. Bookmark the permalink.

4 Responses to Low Altitude Alerts

  1. Jim Oeffinger says:

    So did you ask them what they were showing for your “Mode C” altitude?

  2. Charlie Baker says:

    Mr. Mc Clellan –

    Good commentary on the low altitude alert issue. As a 20,000+ hour retired airline pilot my only negative comment regarding your piece is that you missed a golden opportunity to drive home the point that when receiving such a warning from either ATC or a GPWS in IMC or at night that the pilot must assume the warning is real. As such the prudent and only proper course of action is to immediately establish a maximum angle climb until the warning is resolved. Too many planes and lives have been lost, both civilian and airline, because the pilot wasn’t sure if the warning was valid or not and tried to sort out the discrepancy before reacting to the situation. It’s too easy when you get such a warning to look at your instruments (the ones that you’ve been looking at all along and that may have led you down the Primrose path), see nothing out of the ordinary and assume that it’s “just another bogus warning.” It’s extremely risky to attempt to resolve a discrepancy between where you think you are and where ATC (or your GPWS) thinks you are without taking some action to the warning itself first.

  3. Cary Alburn says:

    We’d all like everything to work perfectly, but the reason we have back up systems is that they don’t always do that. Better a warning than to have “fallen asleep at the switch”. If the warning comes, double check the other basic indicators besides the glideslope indication on the gauge, such as the altimeter as the most obvious. We’re all capable of making mistakes, but if the altimeter shows that the airplane is at an appropriate altitude for that segment of the approach, the glideslope indication is appropriate, then no worries. Not sure that I’d immediately go to a full power climb as Charlie suggested, but I bow to his superior experience.


  4. Bob says:

    Cary, in the moments after the alert that you spend cross checking your instruments to see if you’re at the correct altitude, can you be absolutely certain that you won’t fly into the side of a hill or mountain or tower? It’s a no brainer if you’re VMC, just look out the window, but would you bet your life and those of your passengers on it while in IMC?
    In too many instances, an aircraft has been flown into terrain causing injury and death because pilots elected to try to figure out whats wrong rather than take immediate action and climb to a safe altitude first. The procedure that Charlie describes was borne from history, in an attempt to stop it from repeating itself!

Leave a Reply

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


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>