Squeezing past T-storms: are there simple rules for thunderstorm navigation even a computer could use? Possibly, but success for the human pilot is as much intuition as information.
Flying an aircraft is a process, and complicated manufacturing processes are successfully automated every day. The first step in process automation is to determine the inputs. The second step is to analyze that information and then act upon it. Those actions are the outputs.
We have a long list of inputs when it comes to thunderstorms. The first is a preflight weather briefing, followed closely by our vision. Datalink weather data provides a wealth of data including a composite NEXRAD image. Sferics technology in the form of a Stormscope or Strikefinder provides real-time electrical activity data. Some of us have onboard radar. And ATC or Flight Watch can give us a ground-based perspective on what's happening.
If we could quantify those inputs and resulting actions, all of us would benefit. The problem is that answering the fundamental questions requires as much intuition as science. Those questions are: Where is it, and how close can I get?
Where Is It?
In day VMC conditions, finding the storms is a pretty easy task for a pilot. But as the computer can never see and the pilot can't see in the clouds or at night, we need other sensors to answer the "Where is it?" question.
Much emphasis is placed on the importance of a preflight briefing, but in the automation model of process control, decisions are made based on real-time data. An autonomous aircraft operator or pilot might use a preflight briefing to plan a route, or make a go/no-go decision, but it's real-time information we need to formulate a weather navigation and deviation plan.
Datalink NEXRAD images display an incredibly precise picture of the activity. The problem is, because it's so precise, we tend to believe it's that accurate. Much has been written about the inherent errors in datalink, but the short story is this: Datalink must be confirmed by another source. Thunderstorms can develop quickly and they seldom stand still. Some interpolation is required for both of these dynamics.
If there's a developing line of convective activity on the Datalink NEXRAD image, it's neither safe nor smart to assume that the area you plan to penetrating is free of activity, unless you can see with your eyes, onboard radar or sferics no cells have popped up or moved into since the last datalink update, or aren't heavy enough to show on NEXRAD.
Onboard radar is a great way to confirm that a NEXRAD image is valid, but it has limitations as well. The tiny antennas mounted under the wing of a single-engine airplane have limited range. Learning the tilt angle that will find the ground behind the depicted weather is both an acquired skill and fine art. Being able to discern the difference between a storm cell and a city takes practice, especially on an active weather day.
Heavy rain can absorb all the energy being radiated, instead of reflecting it, and this can mask a monster storm behind what appears only to be heavy rain. Datalink images mitigate these limitations somewhat, but again requires experience to sort the information out of the noise.
Sferics detects lightning bearing relatively well, but range is a rough estimate, at best. Other limitations, like radial spread, require experience to be able to accurately interpret the data being displayed to make it useful. Sferics alone is useful information, but when combined with other sensors, it becomes more meaningful.
A pilot can communicate with ATC or with Flight Watch to get additional information, and a dispatcher, or flight controller of an autonomous aircraft, could communicate with the aircraft and modify its flight plan as needed. But since datalink weather now dumps nearly the same information available to a Flight Service Specialist or ATC into the cockpit, pilots with that system seldom need to call.
The best confirmation comes from the pilots eyes. Even though they are only useful when the pilot can see, they add a nearly irrefutable data point regarding the location of a cell. Our automated aircraft may not have that luxury (maybe some image-processing technology will fill in some day), so count one extra input for the live pilot.
Knowing where the storms are means we can deduce where they are not and fly clear of the threatening weather. The problem is, whether it's an autonomous freighter flying 100 tons of cell phones from Malyasia to Memphis or an anxious father flying to his daughter's wedding, how much clear? Airplanes need to be dependable as well as safe transportation.
How Close Can I Get?
Imagine if the freight companies shut down their operations every time convective activity ventured within 20 miles of their hub airports. This is rote-level learning and too course and rigid for real mission utility. Stay up late and watch Memphis or Louisville on a flight tracking website with a radar overlay. There are nights when the aircraft wind their way down the arrivals in what looks like a slalom course.
Or be at the airport in the morning when the few remaining check haulers, (most of whom have been relegated to buggy-whip status by electronic check clearing, but I digress) or single and light-twin overnight express freighters arrive. Watch them weave through the weather with ease on the flight tracker. Then when they land, ask them how their trip was. Most will respond, "Not too bad, just a few rain showers and some lightning." If these folks adhered religiously to a 20-mile standard, the air transportation system would fold.
If 20 miles from a thunderstorm is not the gold standard, then what is? The answer is, of course, "It depends." All the data is absorbed into a multi-dimensional mental matrix that includes the aforementioned elements combined with other quantifiables like altitude, surface wind speeds, winds aloft, maximum tops, whether the storm is developing or dying, whether it is an isolated cell or part of a line, and how fast it's moving across the ground.
Then come the intangibles. What the storm looks like is important, but the automated aircraft won't have to worry about the other things that factor in for us humans: how tired the pilot is, if this is a familiar area of the country and whether the spouse and kids are on board.
There are times when the textbook 20 miles is not far enough. Pilots flying at FL410 who look up and see an anvil top above them with warts of warm, moist air breaking into the tropopause would do well to yield more than a 20-mile berth to the super cell--especially on the downwind side where hail might be pitched out into miles of clear air. Yet below the bases at 1500 AGL, a five to 10-mile margin might be adequate clearance for a clearly visible rain shaft and result in a perfectly smooth ride.
Whenever possible, thunderstorm flying should be accomplished VFR. The biggest plus is that you have no clearance to comply with and can turn any direction at any time without asking for permission. Flying at 1500 AGL in VMC is a far better option than at 4000 feet in the clouds trying to use technology to avoid embedded cells. Even on a stormy night, being VFR is better. The lightning will highlight the trouble spots.
All the electronic sensors in the world won't replace the ones in our head. Eyes will not only tell the intensity of the rain, seeing the outflow from the storm, a shelf or wall cloud that precedes the storm and the actual width of the holes or breaks is the best way to decide whether to punch or run. If there is a wall or shelf, most likely there will be a gust front, and those bring the possibility of severe turbulence and wind shear.
Other visual attributes of a storm can be analyzed by the meteorologists, but a freight pilot looks at a storm in simple terms. First impressions count. Large, organized lines of weather with power and majesty get respect. Steer well clear of those with green or purple bellies.
When the storms bury themselves in the clouds, then the game gets tougher. Before the advent of datalink weather, the drill was to look at the last NEXRAD composite before take off, formulate a plan, get airborne, call ATC or Flight Service and confirm that the weather was developing or dying in accordance with the plan. From then on, it was whatever could be seen, or pried out of ATC and Flight Service.
Working Blind with Data
Datalink allows a strategic approach to the weather instead of a tactical battle. Often, a small excursion results in a smooth ride for pilot, passengers, and the plane and that is always a fair trade. But sometimes a line of weather must be crossed and that requires backing up the datalink with a real time data source.
One tactic with an airplane with sferics is to fly up close to a break in a line identified with datalink or by ATC. A well-developed line of weather shows solid returns with radial spread all the way back to the airplane on a sferics device. But when you see a hole start to appear and it becomes a pie shaped open area, that confirms the hole. Crossing the line, the sferics device will look like it flipped over as the pie shaped opening becomes a line, and then a pie shaped opening behind the aircraft. This tactic uses the instantaneous sferics information to confirm that a cell is not developing in the hole between updates of the datalink in an active system.
Again, we are back to that tough question: How wide does the hole need to be? There's still no easy answer. Using the cloud tops feature on the datalink will provide another data point, if you have it. That just adds another dimension to the already complicated matrix of data being considered but it is valuable information.
Pilots who have on-board radar can verify the datalink data with the ship's radar. However, cells are often embedded in rain or showers. This makes it hard to tell if the radar is attenuating and there is a monster cell hiding behind the depicted rain.
Backing up radar with datalink or sferics really builds confidence when it comes time to punch. Also, if a hole or break is small, tilting the radar up high before you get close to the line will help center the aircraft between cells. The lower spots appear as holes when tilted up and as a solid line when the antenna is looking straight forward. The objective is to aim for the spot with the lowest tops overhead.
Another thing to remember when using onboard radar is that heavy rain does not always equal thunderstorms or turbulence. Rain reflects radar energy and there is not always a direct correlation between rain and turbulence.
Sferics devices sense the static electricity between moving air masses and lightning. If the radar and/or the datalink shows heavy precipitation but the sferics device is clear, and especially if the datalink displays low cloud tops, chances are good that the ride will be smooth, even though convective activity may be in the area.
No Good Answer
So, no matter how good we get at "where," "how far" will stay unanswered. Pilots routinely and safely fly closer than 20 miles to significant convective activity with passengers or packages. Pilots also sometimes get hard lessons that 20 miles was not enough.
The complicated decision-making process is not linear, and possibly not even logical. It draws on training, knowledge and, most importantly, the experience of the pilot. We have the technology to autonomously fly the aircraft from departure to destination, and determine the location of the bad weather. A missing link in the long chain that will make autonomous freight haulers possible is how to run the slalom course.
Will a few lines of code tell the airplane how close it can fly to the weather? I think the pilots flying purple and brown boxes around the skies have job security for a long time. As for your own decision-making process, you still have the best machine we know of to learn from its successes and mistakes, and refine its logic from its own experience.
Doug Rozendaal is former freight pilot and IFR contributing editor.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||WEATHER SMARTS|
|Date:||Aug 1, 2010|
|Previous Article:||Put down that plate: you've traded a fat bag of money for a bunch of gee-whiz avionics in your panel, but you're still getting all your data from...|
|Next Article:||Drink up after landing: you could fly the ICAO plate for the ILS into Dundee, Scotland with no more than a U.S. ticket and good intentions. That'd be...|