Say Again? #71: Weather Radar

  • E-Mail this Article
  • View Printable Article
  • Text size:

    • A
    • A
    • A

Say Again?

You may think it's the wrong time of year to write about thunderstorms but that is what I've chosen to write about. My reasoning is that there is much to learn -- and much I can't teach you. I'm going to talk about various things but my goal is to get you to do some research on your own. I want you to have time to prepare before the next thunderstorm season.

I worked with "weather radar" my entire career. I suspect I've spent more time watching thunderstorms on radar than virtually any pilot flying. I don't mean to brag; it's just the nature of the job. My point is that, after 25 years, it's amazing how much I don't know about thunderstorms. The biggest point I want you to take away with you is that thunderstorms, for our purposes, are unpredictable. Just when you think you're getting a handle on how they act, a thunderstorm will do something unexpected.

Raw Radar

When I first started out as a controller, we still had the "raw" feed available from our radar sites. You could flip a switch and the display would change from the digitized, mosaic image (called narrowband) to the broadband image. This was the same image that controllers used back in the days of "shrimp boats." Nobody wanted to push those little pieces of plastic around on the radar scope after we didn't have to, but the weather display was still quite useful. There were even a few guys around who still knew how to use the controls at the sector to tune up the image. I never learned how to do that, but I did learn how useful broadband was in displaying thunderstorms. I remember many days spent flipping the switch back and forth between narrowband and broadband: narrowband to work the airplanes, and broadband to get a better picture of the thunderstorms.

I don't want you to think this is just a trip down memory lane. I want you to think about the differences in the radar. The broadband image was from a single site. Like all radar, it had the limitation of not being able to see through, or around, a storm. Narrowband (the digitized image) is a mosaic image. It uses multiple radar sites and it can see on the back side of a thunderstorm (by using another radar site.) Do you see the correlation to today's environment? Your typical airline pilot has his own radar that he can tune to his liking. The typical general aviation pilot has a mosaic image (NEXRAD) that can see the back side of a thunderstorm but the user is at the mercy of whoever "tuned" the radar image. Keep that thought in mind. We'll come back to it later.

Y'All Come

I've been inviting y'all to the Communicating for Safety Conference for years now. I know some of you might think I'm shilling for NATCA but I don't work for NATCA now and I'm still inviting you. Take it for what it's worth. It will be in Atlanta this year and I plan on attending.

At last year's Conference we had one of the most fascinating speakers I've heard yet: Dr. David Strahle. His presentation was, "Basic Principles and In-Flight Use of Datalink NEXRAD Radar." Unfortunately I missed the beginning of the lecture -- a long story I won't bore you with. Anyway, I gather that Dr. Strahle was in on the very beginning of using NEXRAD for in-flight weather avoidance. I got a lot out of the lecture but, as usual, it was something odd that grabbed my attention. Dr. Strahle was relaying a conversation he had with a controller as he was flying around some thunderstorms. The gist of it was, the controller was a lot more nervous than Dr. Strahle was. The controller didn't know what NEXRAD was and he just wasn't used to seeing any GA pilots fly around in weather that bad.

That sentiment fit in with a theory I have. Well, it's more of a guess than a theory. In case you haven't noticed, there has been a sharp increase in the number of GA accidents in thunderstorms. I believe that is due, in large part, to the fact that GA pilots are flying around in weather in which they used to stay home.

Safety Alert

The situation has lead the NTSB to issue a Safety Alert. AOPA has put out its own warning. (Both are Adobe PDF documents.) Both say the same thing about the area I want to concentrate on. AOPA says, "In 2004, nearly 25 percent of all fatal weather-related accidents involved encounters with thunderstorms. Amazingly, in all those accidents, the pilots flew into extreme conditions despite being in contact with Air Traffic Control (ATC). The NTSB's alert says, "These accidents have all involved aircraft operating under instrument flight rules and in contact with air traffic controllers."

As you can imagine, that is somewhat painful for an ex-ATC safety rep to read. I'm going to talk about some of the reasons -- some of the psychology -- behind it, but I'll start off by saying something controversial, just to get your attention: If you are using an air traffic controller as your primary thunderstorm avoidance tool, you're making a mistake. That statement needs to be qualified in about 16 different ways, but I want you to take the statement itself to heart.


First and foremost, you need to understand what type of controller you're dealing with. Is it a Center controller or an Approach controller? If it's an Approach controller, which Approach control? There are about a half-dozen issues to deal with in this area alone. Pilots can't imagine how removed Center controllers are from the rest of aviation. Most of them don't even work near an airport. Unless they have some outside interest in aviation, many don't even know what most airplanes look like, much less what kind of deicing and weather equipment they have or don't have.

As the FAA tries to consolidate more TRACONS into Large TRACONS (like Atlanta), the condition will creep into some Approach Controls. This "removal" from the rest of aviation is subtle but it is also very real. The target of a Cirrus looks just like the target of a 747 on radar. In the rest of aviation, they look nothing alike. In a radar controller's world, they do.

Back to what kind of controller you're working with. Is he a new guy or an old guy? Does he have much experience working with thunderstorms? How does set up his radar scope? Does he trust his equipment? (That one is huge.) Is it set up so he can see the thunderstorms or is it set up so he can see the airplanes? You don't get to ask these questions and you don't get answers to them. But I assure you that every one of them is important to your safety.

It all comes down to this, from the ATC "bible":


The primary purpose of the ATC system is to prevent a collision between aircraft operating in the system and to organize and expedite the flow of traffic.

A controller's primary job is to separate airplanes. He will set up his equipment to accomplish that primary job, even if it means turning the weather display off so he can see the traffic. I cannot overemphasize this point. If he can't accomplish any other task, he will accomplish that one -- separating airplanes. And if he gets that busy, he's not going to stop and tell you that he's too busy to tell you about thunderstorms.

Altitude Squawk

Thunderstorms don't have transponders. "Knock, knock. Hello in there." Just seeing if you're paying attention. Controllers see targets in the middle of thunderstorms every time they see a thunderstorm. They might be over the storm or under it, but it's displayed in the same place on the two-dimensional radar scope. A target headed toward a thunderstorm doesn't ring any more mental alarm bells than anything else. And that is what I want to talk about for a little bit: The mentality -- the psychology if you will -- of working with thunderstorms.

I'll have to stick with the Center aspect of it because that is really all I'm familiar with. In general, you'll get better handling around thunderstorms from Approach controllers. The most obvious reason for that is airspace. Approach controllers only own (and work) the lower altitudes. Let me explain.

ATC Thunderstorm Graphic

If you'll remember way, way back to Say Again? #8: Air Traffic Chaos, I showed you what precipitation looked like on our radar scopes at the Center. I'll need some nomenclature for all these different displays so I'm going to call this the "stick figure" weather display. I'm also going to call it "weather." Just keep in mind it is actually precipitation. We don't paint any weather (or clouds) until there is enough precipitation in it to show up on radar. So the figure at right is a typical stick-figure thunderstorm.

Can you tell me what altitude the first line (_________) is? How about the altitude of the first "H"? The lines (________) represent moderate precipitation. The "H"s represent "Heavy" precipitation. The truth is, controllers don't know what altitude the thunderstorms are active. Hey, I've got an idea! Let's ask a pilot. But which pilot? Well, the one you're working, of course. And here's the kicker if you're a Center controller: Where you're working determines what altitude you're working and, usually, what type of pilot you are working.

"Delta one twenty three, Center, radar indicate moderate precipitation at 12 o'clock and 10 miles. Do you see anything out there?"

"Nothing out here, Center, but blue sky."

Can you see what a huge difference the pilot's perspective would have on his report? That Delta pilot was at FL370. How different would it be if he was at 7,000? How different would it be if he was in a C172? Seriously, try it again with a Skyhawk.

When the controller said "see" to the airline pilot, you were thinking about radar, weren't you? With a Skyhawk, nobody -- no pilot -- is thinking a Skyhawk has radar. But what about a first-year controller? And as soon as you say, "Well everybody knows that Skyhawks don't have radar," some Skyhawk pilot will chime in that he just bought himself a brand-new gizmo with a NEXRAD display.

The NEXt Big Thing

Speaking of NEXRAD, as I told you in Say Again? #36: Spring is Sprung the Centers now have a NEXRAD image displayed on our radar scopes. I personally think it's a great improvement over the stick-figure display and even over the old broadband display. It comes with a couple of really big "buts," though. The display isn't fast enough to be considered real-time. It's close, but not close enough. And as I mentioned earlier, I believe trust is a huge issue in figuring out why we're having so many accidents.

The cold, hard fact is that many, many Center controllers just don't trust their equipment when it comes to displaying weather.

Let's go back to my earlier scenario.

"Delta one twenty three, Center, radar indicate moderate precipitation at 12 o'clock and 10 miles. Do you see anything out there?"

"Nothing out here, Center, but blue sky."

Let's pretend you're a Center controller with a NEXRAD image showing a little blob displayed. Which is right, the NEXRAD radar or the pilot? How about both? Let's think about this.

A plane at FL370 is 7 miles straight up. Add in the 10 miles for the distance and figure out the slant range. Considering the murk that passes for "clear" here in the East, is it any wonder the pilots can't see the thunderstorm? But what about their radar? We all know that radar is "line of sight" so let's think about line of sight. Just as the pilots can't see over the nose, the radar isn't looking straight down. It "looks" out at an angle too. Can you see how an airborne radar could look over a developing thunderstorm?

I've seen this happen. Three airliners all told me nothing was out there. So did the fourth one. But he ran into moderate turbulence right when he went over the thunderstorm. After that, everyone suddenly started "seeing" the thunderstorm.

While I want you to take note of the technical aspects, I want you to look deeper into the psychological aspects of this. I "cried wolf" to three airplanes and no one saw anything. If the story had stopped there, how confident would you be in your NEXRAD presentation if you were a controller? Confident enough to vector a pilot without radar around thunderstorms with it?

Again, let's go back to the Skyhawk:

"Skyhawk one two three, Center, radar indicates moderate precipitation at 12 o'clock and 10 miles. Do you see anything out there?"

Do we really expect him to see anything when the visibility is not even 10 miles? Maybe. It might look a little darker. That's if he's even in VFR conditions. The only way a controller can know that is to ask.

Look! Up In the Sky!

How about up? Up there around 37,000 feet ... you know, 7 miles up. Let's return again to how things work -- thunderstorms and NEXRAD. We all know that thunderstorms start with warm air rising. It rises straight up until it runs into some wind. Your typical line of thunderstorms in the Southeast moves from the northwest to the southeast as the fronts pushes through. It gets even more complicated as a single cell moves to the northeast along the line. The point being, the wind blows. When you're on the edge of a thunderstorm at 7,000 feet, where's the edge at FL370? Now factor in NEXRAD.

A NEXRAD image can be five to 10 minutes old and it gets really complicated when you try to figure out just how old it can be. Or where it came from. Some areas are blessed with coverage from multiple radar sites. Some areas don't have any coverage below certain altitudes (think mountains).

Doppler Weather Radar Coverage

Let's say you're on the backside of a thunderstorm that is moving east and the heavy rain is three minutes east of you. To a controller using NEXRAD, it could look like you're still in the heavy rain. Conversely, on the other side of the same thunderstorm, he could be showing an aircraft that is in the clear but is actually 3 minutes into the heavy rain. Can you see how a controller gets to where he doesn't believe his weather display?

There's one more piece of the puzzle I want to throw in. It's a quote from "Frequently Asked Questions" page from the National Weather Service's Web site.

Doppler Radar FAQ

Composite reflectivity displays the highest reflectivity of all elevations scans. So, if heavier precipitation is higher in the atmosphere over an area of lighter precipitation (the heavier rain that has yet to reach the ground), the composite reflectivity image will display the stronger dBZ level.

This occurs often with severe thunderstorms. The updraft, which feeds the thunderstorm with moist air, is strong enough to keep a large amount of water aloft. Once the updraft can no longer support the weight of suspended water then the rain intensity at the surface increases as the rain falls from the cloud.

(The link is there so you can read the whole page in context.)

Remember back at the Skyhawk scenario I asked, "How about up?" If you use NEXRAD in your airplane you might want to think about that when your NEXRAD says "bad" but your eyes say "not so bad." What happens when that "large amount of water aloft" runs out of lift?

Comfort Zone

Getting back to my point, I can tell you from experience, it's just another situation that appears confusing to a Center controller and degrades his trust in the weather display. NEXRAD for a controller is a two-dimensional representation that updates slowly -- independent of his traffic display. Think about that for a moment, too. His traffic radar updates every 10 seconds and he knows it is correct. His NEXRAD updates every ... he doesn't even know. And it appears incorrect on numerous occasions. By the way, it isn't even NEXRAD. It's WARP (Weather And Radar Processor). Does this look like your NEXRAD image?


I can't tell you how much the FAA manipulates the data before it reaches the controller's scope but they change the colors ... at a minimum.

Having detailed all these problems, you might have forgotten that I said I liked NEXRAD better than the old "stick-figure" presentation. It helps to remember that I was a safety rep and very experienced before NEXRAD/WARP was deployed. I studied NEXRAD when it came out more than most controllers and I had the experience to compare and interpret what I was seeing. Which is the reason I didn't depend on it alone.

I ran my scope with WARP looking from the ground up. (Sorry, I didn't have room to explain it has altitude-filter limits.) I also never ran it alone. I always turned on the "stick figure" display too. The clutter of so much information bothered a lot of controllers but I found the fast update of the narrowband display filled in a lot of gaps inherent in NEXRAD. But the biggest difference between me and the way most controllers operated was that I vectored for VFR conditions. If you wanted me to vector you around a thunderstorm, I vectored you around it. For general aviation airplanes, that meant a vector out of all the rain displayed and that almost always put them in VFR conditions. It's amazing how your comfort level goes up when you can see. Yours and mine. No one ever complained. And I never lost one.

Have a safe flight.

Want to read more from Don Brown? Check out the rest of his Say Again? columns.