Weather forecasts – there’s more to it than just charts

In scheduled air carrier operations, it is a well-known tenet that you cannot depart if the weather conditions at your destination will be below your authorized landing minimums when you get there. This requirement goes back as far as 1937, when Civil Aeronautics Rule 61.71090 said that:

“No scheduled airline aircraft shall be dispatched unless the observed weather information and current weather forecasts, pertaining to the point cleared to, give sufficient indication at the time of clearance that the ceiling and visibility are or will be, when the flight would arrive at such point, at or above the minimums specified in the weather competency letter for letting-down-through.”

approach fog
How low is too low?

This seems like a prudent idea, particularly when you consider that only 11 years earlier, while flying the mail, Lindbergh himself had parachuted out of not one, but two airplanes when he ran out of gas before finding someplace with enough visibility to land. With the advent of the Civil Air Regulations, one has to wonder whether the image of a revered aviator such as Lindbergh taking off… twice… when he had no earthly idea whether he would be able to land might have motivated the regulatory language. In any event, today FAR 121.613 uses similar, but critically different, language:

“…no person may dispatch or release an aircraft for operations under IFR or over-the-top, unless appropriate weather reports or forecasts, or any combination thereof, indicate that the weather conditions will be at or above the authorized minimums at the estimated time of arrival at the airport or airports to which dispatched or released.”

Language regarding a “combination thereof” was added in 1955. In 1963, the phrase “weather reports and forecasts” was changed to read “weather reports OR forecasts.” In 1967, the phrase “ceilings and visibilities” was altered to read “weather conditions,” primarily to remove the requirement for a ceiling when the approach in question did not use ceiling as a minimum.

The FAA still relies on a legal interpretation issued in 1989. The key paragraph in this opinion says the following:

“The rationale behind the current regulation is that as long as one can show a combination of weather reports or forecasts indicating above minimum weather conditions at the estimated time of arrival (ETA) at the destination airport, the flight may be dispatched or released. However, the converse is also true, that when any combination of weather reports or forecasts show below minimum weather conditions at the destination airport at ETA, the aircraft may not be dispatched or released.”

In today’s scheduled operations, this is normally not an issue. Contemporary Part 121 operations are heavily influenced by Exemption 3585, which generally allows departure even if a conditional portion of the forecast includes weather conditions as low as one-half the required minima. Exemption 3585, however, merely changes the numbers and adds a second alternate requirement; it does not change the 1989 interpretation of just what 121.613 is trying to achieve.

On two recent occasions, I have spent my day staring down FAR 121.613. Both cases required a more in-depth study of the day’s weather than a simple scan of the TAF.

In the first situation, we were scheduled to operate from Dallas to Albuquerque, which anyone in my MD-80 fleet has done a hundred times. On that day, however, the ABQ forecast called for winds out of the southwest at between 20 and 30 knots, and a broken cloud layer at 3500 feet. Of course, there are no straight-in approaches to the west runways at ABQ; unless you are skilled in space shuttle operations, the mountains make straight in arrivals sort of problematic. There are a couple of RNAV approaches, however the airplane is precluded from flying RNAV instrument approaches at ABQ.

Normally this kind of wind would be handled by flying the RNAV Visual approach to runway 26. However, to do that you need visual contact with the airport at around 10,000 feet as you cross the Sandia Mountains; the broken ceiling wasn’t going to allow that. The obvious solution was to fly an approach to runway 8, then circle to land on runway 21. Because my operator does not have a training program for circling approaches, our Op Spec limits us to a ceiling of 1000 feet and three miles visibility. The forecast met those minima easily.

Forecast Discussion
The Forecast Discussion often contains some valuable details about weather trends.

However, the forecaster’s written discussion for that day, contained in the National Weather Service Area Forecast Discussion, described in detail the major winter storm that he expected to churn across the New Mexico mountains, which included travel advisories for all routes leading east and north of the city. (We have to be careful of this terminology; the NWS Area Forecast Discussion contains an Aviation Weather paragraph. This has nothing to do with the traditional area forecast which we know as the FA, which is on the verge of extinction; rather, the NWS aviation paragraph within the Area Forecast Discussion is described in AC 00-45H, paragraph 5.20.)

In this case, the forecaster also said that there was some uncertainty as to whether the snow would impact the downtown area, although the current TAF did not include any mention of that. What I needed to keep an eye on was what kind of ceiling and visibility was actually going to develop during the day, and how that might impact the circling maneuver. Two questions were open: would we be legal to depart on the basis of the circling approach, and would we actually do it? Not the same question; I’m not inclined to circle an airplane weighing 130,000 pounds, at night, out of a strong tailwind, at 999 feet with only three miles of visibility… in snow. But with a higher ceiling and better visibility, we could do it.

In the end, the city and airport were wide open, the wind at the surface died off, and we flew a standard ILS to a landing on runway 8. But throughout the day, we had to keep in mind the possibility that, due to the language of 121.613 and the limited approach options, we might not be able to depart.

The second situation involved a strong inversion that was camped atop the entire Midwest. An arctic air mass had slipped under much warmer air, extending down to the south of Dallas while only about 2000 feet deep. Climbing through that 2000 foot level, the temperature warmed by 10 to 20 degrees Celsius, depending on how far north you were. The boundary for this inversion was a lazy front draped between Dallas and San Antonio. The air below the inversion had been saturated with drizzle and rain during the day before; to the north, much of this had been freezing rain. This led to dense fog, with visibilities at DFW down to 1/8th of a mile.

As we perused the METARs and the TAF while at the gate in St. Louis, we had to factor in a rather unwelcome variable. While the MD-80 is capable of Category III autolandings, our particular airplane was not; due to a mechanical discrepancy, it had been downgraded to category I operations under the MEL.

Surface analysis chart
Not a pretty picture – but what do the TAFs say?

The front was supposed to move north during the day. The TAF earlier that morning had forecast three miles of visibility by the time our scheduled arrival from STL that evening. Successive forecasts had pushed the visibility improvement back as the day wore on. Nonetheless, the TAF we were looking at in the cockpit prior to departure still called for two miles visibility in the vicinity of thundershowers by our ETA.

The problem was the METARs weren’t obeying the TAF. The visibility at DFW hovered around ¼ mile all day long; as we contemplated departure, the RVR was 1400 variable to 1600… well below our Category I minimums. Which combination of weather reports or forecasts should we use to comply with 121.613?

Again, there is more to the “weather reports or forecasts” language than just the TAFs and METARs. The forecaster’s discussion that morning had said that:

“An attendant surface low will deepen across West TX by midday, and strong southerly flow will ensue in the vicinity of the stalled frontal boundary. As a result, the stalled front should finally lift back to the north beginning later this morning and into the afternoon. Surface winds will veer to the east and eventually southeast and the warm sector airmass will begin to overspread North and Central TX.”

We could see, from the METARs, that the wind was indeed veering and the temps and dewpoints were rising. But was the front moving fast enough? The forecaster had also described two waves of convective activity that would be boosted in development by a low pressure system entering west Texas; and indeed, radar showed a line of storms popping up to the west of DFW, which would be the first wave that had been expected. These indicators made a good argument that the front was moving, and the developing energy would keep it moving. The probability that the latest forecast was accurate seemed pretty strong.

After a discussion with the dispatcher, we agreed that we were good to go under the language of 613. We departed, and, sure enough, during the final 25 minutes that it took to fly the arrival, the visibility improved from 1/2 mile to 1 and 1/2 miles. The thunderstorms to the west were clearly evident on radar, as were several high based cumulus buildups to the east along our arrival path.

Those high-based cu’s created quite a bit of anxiety during the arrival for several pilots. We had the advantage of being at the front of the pack and already in the descent; we both agreed that we would be well below the bases as we passed by them. But another statement by the forecaster that morning gave me more confidence in this assessment; he had noted that:

“While much of this buoyancy will be confined to the mid levels where cooling due to ascent from the approaching upper low will steepen lapse rates, the warm/moist air at the surface may result in surface based instability. If an isolated storm is able to become surface based during this time, there will be a severe potential including tornadoes given the ample shear.”

The notion that much of buoyancy would be confined to the mid levels was consistent with the inversion that had been sitting there for days, and this seemed to be borne out as we flew beneath a widespread but quite benign stack of cumulus.

Radar tornado TX
An isolated storm became surface based – and the result was ugly.

But the forecaster wasn’t done; a couple of hours later, while sitting in the cockpit with a fully loaded cabin at departure time, we were stuck at the gate when the DFW ramp closed due to lightning. Looking at the radar, we were suitably impressed by a very, very red cell to the south of the airport exhibiting a well-defined hook. Apparently, an isolated storm was able to become surface based… and a few minutes later, we expeditiously deplaned everyone when a tornado touched down a couple of miles south of the field.

The Aviation Weather paragraph, described in AC 00-45H, can be found at the Aviation Weather Center website. Under the “Forecast” pull down menu, you will find the phrase “Avn. Forecast Disc.” This will open a map, allowing you to click on each forecast area to pull up the aviation weather discussion.

However, the excerpts from the forecast discussions that I cited above did not come from the Aviation Weather paragraph, but rather from the near term, short term, and previous discussions, and updates, that are posted by the forecaster in the complete area forecast discussion. The entire list of complete discussions from each forecast office is currently available at the National Weather Service Central Region Headquarters website.

You can also find them on the local webpage for each National Weather Service forecast office. This may take a little hunting; different offices seem to employ different webpage designers.

Regardless of which part of the FARs you are operating under, the area forecast discussions put out by local forecasters are incredibly valuable when preparing for a day’s flying. They will give you the feel of a personal briefing; my favorite was one a couple of summers back, on a very complicated convective day, when the forecaster started out with the phrase, “To convect or not to convect; that is the question.”

The information contained within these discussions can often give you a heads up about conditions not yet shown in the TAFs, and the overall picture you can develop will give you benchmarks with which to decide if the day is developing the way the forecaster thought it would. And at the risk of sounding old-fashioned, all of the automation notwithstanding, weather interpretation requires a human interface. While a picture is worth a thousand words… and good weather graphics are… a couple of hundred words authored by a sentient being can complete the picture nicely.

Leave a Reply

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