Safety Pilot: Weather wolves II

October 1, 2008

Bruce Landsberg has led the AOPA Air Safety Foundation since May 1992.

Cry “wolf” often enough and people stop listening, even when the wolf may really be there. It’s the same with many of the weather warnings given to general aviation pilots. Ever had a forecast that substantially over warned of a particular hazard? It’s a rhetorical question—we all have. And it’s tough to make good decisions with bad information but like so many complex issues, there are shades of gray.

I wrote on this topic years ago, November 1995 to be precise, and some Flight Service briefers took exception since they were just doing what they’d been told ( “Safety Pilot: Weather Wolves”). Weather capriciousness, FSS policy, and perhaps an unwillingness to call the weather as it was just didn’t jibe. The system has not improved noticeably but now the FAA and Lockheed Martin, which took over the FSS function, can share the blame. To be fair, Lockheed, again, is only following the FAA requirements of its contract but there are some things that could be done better and I will send these comments on to the usual suspects.

The most overused warning is “VFR not recommended,” or VNR. This is issued any time the ceiling is forecast to be below 1,000 feet or the visibility less than three miles. That seems reasonable until we look at the boundaries of the warning. It covers at least a 3,000-square-mile area (significantly bigger than Delaware or more than half the state of Connecticut) for a period of six hours. The FAA requires a large area and time interval even though the National Weather Service has the ability to get considerably more specific. This is the requirement for an airmet and those requirements were established decades ago before the technical tools were as good as they are today.

So, suppose you’d planned a flight at 11 a.m. departing from “Inland airport” but the forecaster was thinking about low clouds along the coast 50 miles away in the early morning at 7 a.m. Under the current system, the briefers must tell you VNR even though the fog may have burned off hours ago and/or is not on your route of flight because it falls within the area and time period covered by the forecast.

A recent example involved an experienced AOPA Air Safety Foundation staff pilot who was flying VFR to Oshkosh in the early morning. VNR was given even though the affected reporting stations were deep in Appalachian river valleys where ground fog was expected for perhaps an hour or so after sunrise. At his cruise altitude there wasn’t a cloud over half the continent. This wasn’t even remotely a safety issue.

Over-warning leads new pilots to cancel flights that would have been perfectly safe to fly. After being burned a number of times they start to “take a look.” That works well enough and these pilots build up a happy set of experiences based on a certain set of conditions. Later, these folks begin to routinely ignore the “wolf” cry and may eventually launch into truly bad conditions with occasionally fatal consequences. Who’s at fault? Depends on how good your survivors’ attorney is.

There are lawsuits the government lost because a forecaster didn’t include a small area in the airmet and the pilot blundered into cloud or ice and lost control. And you wonder why Uncle Sam is conservative? Over-warning seldom results in a successful lawsuit but failure-to-warn almost always produces large damage awards, regardless of culpability. What’s bad about this state of affairs is that the forecast instruments become exceedingly blunt, not for technical reasons but for wealth transfer avoidance.

I’d have probably sided with the government on many of these cases because weather forecasts are just that—not guarantees. They can’t assure you of bad or good weather. Forecasts are probabilities and sometimes the percentages are close to 50-50. While we all love to blame the weather service, prognostication is not easy. Anyone who’s ever planned a picnic can vouch for forecast fickleness. But then barbecue miscues don’t usually result in a high-speed impact with the ground unless the cook has gotten into the beer cooler, but I digress.

Clouds do form out of thin air but there is generally ample warning as other clouds appear above or below the flight altitude or a close temperature/dew point spread foretells of saturated air. If you slept through that section of ground school or home study, better go back to review ASF’s online course on Ceiling and Visibility . We lose somewhere between 20 and 30 aircraft a year to VFR into instrument meteorological conditions.

My beef, despite the system being overly conservative, is often there is no interpretation. Some of the briefers are not briefers—they are readers. There is no attempt to explain what might be happening if it’s contrary to the written word. Why?

Here’s a personal example of another non-briefing from FSS although it wasn’t for VNR. On a northbound trip from Florida this spring there was a cluster of thunderstorms around the destination airport and along the inbound route. I landed 100 miles short to let the excitement pass. At the FBO there were five-minute Nexrad weather radar updates, probably identical to what FSS was seeing. After 90 minutes the weather had all moved east except for one small area that was 75 miles to the west but moving slowly eastbound. It would have been hours, if at all, before it was operationally significant.

Upon filing my IFR flight plan to complete the trip, the briefer advised of the convective sigmet that was still in effect for the route and destination. What should have happened next, in my view, was he would then have said something such as, “Your route is presently not showing any activity. It may redevelop later so be advised.” That would have presented a true picture. Instead he was incredulous that I would even consider launching in face of the convective sigmet. Twenty miles east of my present position the concern was justified. I understand briefers’ frustrations with high-risk-tolerance pilots who always end the conversation with, “We’ll go take a look.” Most of them make it.

You’ve all heard my diatribe on the importance of pireps, which provide both input to the system to update forecasts and let other pilots know what’s actually happening out there. As we get into VNR and icing season this is critical. If all this were easy to solve it would have happened by now. In any case, I’m going to propose a novel idea that has absolutely no chance of adoption. How about government truth in forecasting and briefing while pilots take responsibility for not getting into weather that neither they nor their aircraft are ready to handle?