You go first

We had been watching the weather from before takeoff and as often as possible during the 45-minute flight to Butler County Airport (HAO) in southwestern Ohio. There were several good weather alternates close by so safety was never a concern, but the odds of making it to our passengers’ desired airport seemed bleak as the weather reports were constantly far below ILS approach minimums.

Since it was early morning, the controllers had no information to pass along from other pilots ahead of us. We would be first. If we had been flying under for-hire rules the only options would have been to hold and wait for improvement or divert. But the federal aviation regulations allow those of us flying for our own reasons under FAR Part 91 to fly an instrument approach with no landing airport weather information, or with a report below approach minimums. And so we did.

It was an unusual morning because just as we descended on the glideslope inside the final approach fix we had a clear view of the ground below and ahead. Shortly thereafter we could clearly see the runway.

And what we could also soon see was that a dense fog bank covered the southern edge of the airport leaving the runway in the clear. Under that fog was the automated weather observing system (AWOS). Its report was accurate. The visibility at the AWOS sensors really was near zero, and the ceiling, too. It was both a perfect example of the wonderful addition to flight safety the automated weather observing systems have been, and also their limitations.

Before automated weather sensors were first deployed in the early 1990s, only a handful of airports in the United States had weather observations and reporting. The observations were made by a trained human using sensors such as anemometers, thermometers, ceilometers, radars, and a wet bulb thermometer. And they also used their eyes and experience to look around at the sky in all directions and report what they saw.

The human generated “sequence reports,” as they were called, were the gold standard. But these observations were few and could be located 100 or so miles away from the nearest observation point, leaving giant gaps in our information of conditions at the dozens of airports between. Those are the very airports that make our personal and business airplanes so valuable.

Something as basic as the altimeter setting was often only available from an observer many miles away. That meant approach minimums were high because there was no way to know the actual barometric pressure at the airport you were approaching. And that’s where the automated weather observation began. The first automated systems broadcast a reliable altimeter setting on an airport that lacked a control tower or qualified weather observer.

Soon sensors were developed to automatically track wind direction, speed, and air temperature. That was information we just didn’t have before AWOS and it was a godsend. But what was missing, and what we really wanted to know, was the cloud cover, ceiling height, visibility and presence of precipitation, and what type of precipitation was falling. That was the hard part for automated sensors and required years to develop and deploy.

There are around 1,000 automated weather observation systems operating now and they are under slow, but continuous upgrade of the sensors. The systems are named AWOS or ASOS depending on their place in the national weather data collection system. But that makes little difference to us as pilots. Either system observes and reports at least the weather basics.

What we really need to keep in mind is that the automated systems are myopic. They can only look up for a rather narrow view of the sky overhead. And the sensors only measure visibility over a limited range around the system. For safety’s sake the automated sensor array must be well clear of runways which means when conditions are variable the visibility and ceiling height over the runway may be vastly different than observed by the automated system.

Layered cloud cover can be a particular challenge to observe for the automated sensors. For example, if there are only a “few” clouds over the sensors, we would think no problem for a visual approach. But if the cloud layer is broken or overcast a mile away over the final approach course you’d be in solid IFR. And as for reporting cloud cover over the whole traffic pattern, not just the airport, forget it.

Observing and reporting thunderstorm activity is also difficult for the automated systems. Humans are pretty good at hearing thunder—although most dogs seem to be much better—but for an automatic listening sensor to pick thunder out of the continuous background noise of vehicles, airplanes, and the rest? Maybe, but not yet. The automatic report relies on detection of lightning, which is a positive indication of convective activity but difficult to locate with precision, so the AWOS report of lightning is a general direction from the airport.

The FAA and NOAA have continuous programs to update the sensors with ever more precise and reliable equipment. For example, the anemometers are now being heated to shed ice. The precipitation sensors and software are ever better at determining if it’s rain, snow, sleet, or mixed precipitation that is falling. Even so, the system can sometimes fall back to the report of “UP” for unknown precipitation.

The best and most useful weather reports we can get when conditions are close to minimums is an automated METAR that is expanded by human observation. I see this often at my home airport where the tower controllers almost always add a “tower visibility” observation to the automatic look. They can also add a comment on pirep type and intensity or changing weather that is approaching the airport but not yet over the automatic observation sensors.

No matter who or what is reporting the weather, the thing that matters to us as pilots is what we see when we look out at decision height or the missed approach point. The automated reports can tell us to relax while en route, that an easy approach is ahead. Or they can give us lots to fret about like my recent landing in one-eighth mile viz. Bottom line, that’s why it’s called “decision height,” and that’s why the rules allow Part 91 pilots to take a look and make that decision.

J. Mac McClellan is a corporate pilot with more than 12,000 hours and a retired aviation magazine editor living in Grand Haven, Michigan.