Like it or not, ASOS (automated surface observing system) weather observations are here to stay. The administrative and political decisions have been made; AOPA made its views known; and, in spite of so many dissenting views, ASOSs are being commissioned at a rate of about one installation per week. Ultimately, 900 airports will have automated surface observations — all geared to report in the METAR format. Now it's up to us to understand this new way of reporting weather and learn how to apply it to our in-flight decision making.
By the way, ASOS is not AWOS. AWOS (automated weather observing system) is primarily a state-funded effort, and most AWOS units aren't capable of reporting as many weather variables as can ASOS. For the most part, AWOSs will remain in place.
First, the good news about ASOS. ASOS reports of wind speed and direction, altimeter setting, and temperature have proven very, very accurate and reliable. That's because these observations are made directly, by proven instruments. Everyone is familiar with an anemometer's wind vane and rotating cups, the barometer's expandable aneroid cells, and the trusty old thermometer. So when you tune in an ASOS while in your airplane or look at an ASOS-inspired METAR, you can pretty much count on the accuracy of these direct observations.
Wind observations are updated at 5-second intervals, then averaged over 2 minutes' worth of samplings. ASOS also looks for gusts — an exceedance of 10 minutes' worth of sampling by a value of 5 knots or more — and records this value as a gust. The equipment is designed so that the higher the gust speed, the faster the reporting response time. If it's gusting to 30 or 40 knots for 5 seconds or more, for example, the lag time between observation and report would be just a few seconds.
Now the not-so-good news. Because of ASOS's time-based sampling logic and its equipment limitations, great discrepancies between human and ASOS observations have been widely recorded. Trained weather observers routinely see faulty ASOS observations of cloud height, sky cover, visibility, and precipitation type. Moreover, observers often see ASOS crank out various off-the-wall reports that bear no resemblance whatsoever to current conditions.
ASOS uses a laser beam ceilometer to measure cloud heights. It's a narrow beam pointed straight up. Human observers take in the entire sky to help make judgments about cloud heights, ceilings, and sky cover. They can also identify any dangerous cloud types. To ASOS, a cloud base is a cloud base, whether it's caused by a solitary fair-weather cumulus or a raging cumulonimbus.
The ceilometer also uses time averaging in its logic. It makes observations every 30 seconds and double-weights its reports based on the last 10 minutes' worth of cloud base readings. This means that it may take ASOS 10 minutes to report a sudden change in cloud height or sky cover — something to bear in mind in marginal VFR weather or when weather conditions are changing rapidly. Before the 10 minutes has elapsed, the ceilometer may look up into a break in a broken layer and report a clear sky. ASOS reports sky cover in terms of "clear," "few," "scattered," "broken," and "overcast" conditions.
Another problem with the ceilometer has to do with false low-ceiling reports when it's raining or snowing. Precipitation can act like chaff, making the laser signal bounce back off it instead of any cloud bases.
ASOS measures visibility with a forward-scatter meter. A Xenon flash lamp projects its beam to a receiving unit, which measures the amount of light that is scattered by any intervening moisture or other particles. In theory, it's a bit like measuring the headlights of an approaching automobile on a foggy day. That's forward scatter. A back-scatter measurement, on the other hand, would be like judging visibility into the headlight beams of your own automobile as it moves through the fog. We're all familiar with the way this kind of back-scattered light can ruin visibility and even blind a driver on a dark night.
One problem with ASOS visibility reports is that the sampling area is very small: the distance between the bulb and the receiver is less than a foot. This makes siting and localized ground fog a real issue. An ASOS located in a low area that's susceptible to ground fog may be the only place on an airport with reduced visibility. There have been cases where runways and the sky overhead an airport were in the clear, yet the ASOS reported visibilities of one-quarter mile — because the ASOS site was surrounded by a lonely patch of ground fog.
Brightness can also affect the accuracy of visibility reports in certain reduced-visibility situations. That's because the scatter meter is designed to read forward scatter, and we humans perceive visibility through back scatter. (Just remember that when driving in fog you can nearly always see an oncoming car's headlights farther away than you can see into the view ahead of your own headlights).
Let's say that a thin fog or haze has reduced visibilities to 2 miles, according to a trained human weather observer. Research has shown that if it's bright enough, ASOS can report visibilities approximately twice what the human eye perceives. So you hear a 4-mile visibility report on the ASOS frequency. This can trick a pilot into thinking that conditions are better than they really are and spring a potentially lethal trap.
One National Weather Service (NWS) publication says that if it's bright enough to wear sunglasses and there's haze or thin fog, halve the ASOS visibility report. In dense fog, ASOS reports correlate well with human observations.
Visibility measurements are made every minute, then averaged over a 10-minute period. If visibility suddenly drops, say, from 7 miles to 1 mile, it can take about 3 minutes for the 10-minute mean values to register a 3-mile visibility and then make the ASOS transmit a special observation — marked by a SPECI prefix on METARs. It will be 10 minutes, however, before the 1-mile visibility is reported.
ASOS reports vertical visibilities in 100-foot increments whenever the following three events take place: cloud heights drop to or below 2,000 feet agl; the ceilometer beam is "smeared" by returns (presumably) off moisture; and visibility measurements drop to one mile or lower. Vertical visibility reports are signified by the "VV" prefix. METARs have done away with the term obscuration, so don't look for it anymore. Now it's vertical visibility, and ASOS can't tell you if the restrictions to visibility are from haze, fog, dust, or anything else.
Which brings us to ASOS's precipitation identification hardware. This is a device that uses a light-emitting diode to measure the fall rates of various types of precipitation. Snow falls slowly, this instrument's logic says; rain falls faster, and reports are generated correspondingly.
This instrument has a hard time sensing light rain, light snow, and blowing snow. That's because its threshold is set at a precipitation fall rate of 0.01 inch per hour, and lighter precipitation won't hit this triggering level. So light precipitation and drizzle go unreported. Also, the precipitation identifier can report ice pellets or hail as rain because these phenomena have fall rates similar to those of rain. The NWS is working on a fix for this. It involves installing a small microphone on the ASOS. If the unit reports rain and the microphone picks up the racket made by the ice pellets at the same time, an algorithm will tell the ASOS to report ice pellets.
There are other deficiencies in other ASOS components. Lightning detectors have been installed at 26 ASOS sites, and they operate much like the equipment many of us use in our airplanes. They report nearby lightning activity but don't give its bearing from the site. Plans are under way to link the nation's ground-based lightning detection network into ASOS observations. Under this scheme, strike locations would be plotted according to latitude and longitude, then sent to the ASOSs nearest to them, where ranges and bearings from the site could be provided to pilots and weather observers.
ASOS's freezing rain detectors are of the vibrating-probe type used in aircraft icing detectors. When freezing rain adheres to the probe, the vibration slows down and the sensor issues a freezing rain report. Unfortunately, anything adhering to the probe will do the same thing — frost, spider webs, bird droppings, and so on.
I know that all of this may sound like very bad news. However, both the NWS and the FAA are aware of the problems mentioned above — and more — and are working on fixes. Besides, there are advantages to ASOS. These units work 24 hours a day, 7 days a week; don't take vacation; don't go on strike; work better at night than humans; and don't call in sick (unless some elements are malfunctioning, that is). As with AWOS, we can often benefit from lower minimums at outlying airports not served by full-time weather observers, thanks simply to the altimeter setting provided by ASOS — not to mention the rest of the weather information. We have more data more often, and this helps to boost both our forecasting ability and the research community in the quest for better weather prediction. It's ASOS's time-based observation logic, together with the sheer density of observations, that tilts ASOS's accuracy over that of humans in the long run.
What about the fate of human observers? Don't worry; they'll still be busy. And there are no plans for a drones-alone future where all observers end up flipping burgers. Of the 900-odd ASOSs, 400 will indeed be working by their lonesome (they'll be recognizeable by the "AUTO" prefix on a METAR) on what the FAA calls Level D sites, which will most often be small, uncontrolled airports. At the 250 Level C sites, which have part-time towers, observers will augment — that is, monitor and correct, if necessary — ASOS observations. An augmented observation can be identified by the "A" suffix in a METAR identifier.
At the 57 sites with Level B service, full-time augmentation by observers will be the norm, along with runway visual range measurements when visibilities drop to IFR levels, and location of any nearby thunderstorms or lightning. At the 78 Level A sites, cloud type information, variable sky conditions, and sector visibility will be added to augmented ASOS reports.
In fact, many observers have told me that ASOS has proven to be anything but the job-killer it was originally feared to be. "I spend more time correcting than I do observing," a weather observer at a major airport told me. "It's like ASOS is a full employment act for observers … because it's wrong so often." Observers say that ASOS is usually very accurate when stable VFR weather prevails. But when the weather changes rapidly, as it does in frontal situations or when convective weather crops up, be prepared for erroneous information.
The biggest single piece of advice for pilots approaching airports with ASOS: Listen to the broadcast for at least 10 minutes — or at 10-minute intervals — if the weather is marginal, if instrument meteorological conditions are factors, or if the weather is changing. If the readings are all over the place, don't trust them. If they seem to be consistent, then the time-averaging logic is working as advertised and you can expect the actual conditions to correspond roughly with those being reported. And always be ready to go around or perform a missed-approach procedure.
Related Articles
