Wednesday, July 10, 2013

History & Background On Weather Radar

One of the bigger innovations in the weather industry over the past few decades, and certainly the biggest innovation in severe weather forecasting, is the implementation of weather radar. But unfortunately, the discussion on Doppler Shift and Nyquist Velocity and all the other interesting tidbits about weather radar can get a bit boring. But if you promise to read it anyway, I promise to (try to) keep you entertained, and to present you with a much more fun and image-packed Part 2. Sound fair?

Weather radar was first implemented in 1957, with the Weather Bureau’s system of WSR-57s, so named because they were Weather Service Radar from, you guessed it, 1957. Despite the “57,” however, there is always some delay is finalizing new technology. The first of these radars didn’t go operational until 1959, but once they were installed, they remained in use until the early 1990s in most places.

Unlike today’s fancy radar, these early radars were only capable of plotting color for where anything returned the pulse – there was no discretion for intensity, nor was there even a base map. Meteorologists interpreting radar had to place a piece of paper over the screen and trace the storm with a grease pencil, then do so again when the next scan came in, just to have a guess at which way and how fast a storm was moving. They used rulers and measured the distance on paper, and then scaled inches to the actual distance.

Early weather radar sometimes returned sights that were both eerie and clear, like Hurricane Carla in 1961 (courtesy NOAA).

But most of the time, the picture was more muddled. The image below, also from NOAA, shows the Wichita radar scanning a tornado hitting Andover, Kan. in 1991. Unfortunately, ground clutter (which we’ll talk about later on in Part 2) obscured the hook echo.

Wait, did I just say 1991? As in . . . not that long ago?

Yes. WSR-57s remained in everyday use until the mid 1990s, at which time a new fleet of weather radar, termed WSR-88D, took over.

The additional D stood for “Doppler,” and meant that these new radars could measure the speed and direction of the things (usually raindrops, sometimes birds or other debris) being picked up by the radar.

Today’s network of Doppler radars just finished undergoing a massive upgrade to what is called dual-polarization. Dual-pol radars, as they are more commonly called, scan both horizontally and vertically.

(Image courtesy of Wikipedia.)

The original WSR-88Ds produced flat* images of the atmosphere at a certain number of “tilts,” where “tilt” refers to the angle off the ground. Since these flat images are also circles, you can think of them like plates.

(Except that you can’t really think of them like plates, because they’re not really flat, hence the asterisk. As the earth curves below the radar beam, the radar beam actually increases in elevation relative to the earth’s surface, as given in feet AGL, or Above Ground Level. So they’re really more like bowls, and the distance AGL that the radar beam is scanning increases with distance from the radar.)

But back to Dual-pol. Instead of just producing plates/bowls, Dual-pol radar can produce volume scans that show the actual real entire thunderstorm in three dimensions. Additionally, this new type of radar can give an estimate of the size and shape of whatever it’s seeing, so that forecasters can make an educated guess at whether it’s rain or hail, or sometimes lofted debris by a tornado, or whether it’s rain or sleet or snow in winter.

But despite all of this new technology, weather radar will always have a few drawbacks.

One of the most well-known is the “cone of silence.” Every weather radar has one. Weather radar cannot scan vertically upwards, it can only scan horizontally. Even with a high tilt, like 35 degrees, there is still some area very close to the radar that can’t be sampled, because the beam can’t get high enough.

This image, from a Capital Weather Gang article, highlights this phenomenon. Not only is there an empty circle right around the radar site, but the radar returns right next to the circle are distorted linearly, and may not be giving a clear picture of what’s actually happening there.

That’s where TDWRs come in. Most large airports have their own weather radars, known as Terminal Doppler Weather Radars. These complement the WSR-88Ds by scanning in the area that the NWS Doppler radars otherwise couldn’t see.

These airport radars also scan much more frequently – up to once every minute – whereas WSR-88Ds can only scan as fast as every four minutes. They also provide a higher-resolution picture, allowing meteorologists to notice tornadic circulations that would otherwise be hard to pick out.

So if TDWRs are so great, why is the NWS still using WSR-88Ds? The TDWR’s one flaw, and it’s a big one, is that they can’t “see” very far. TPHL, the TDWR for Philadelphia International Airport, is located near Camden, but it can only scan as far southeast as about Atlantic City, and as far northwest as about Pottstown.

Contrast this with KDIX, the Philadelphia-area WSR-88D, which from its base at Fort Dix in the Pine Barrens can scan all the way from central Massachusetts to almost Virginia Beach (not that the data collected on the fringes of a radar are any good, but you get the point). NWS meteorologists need this type of radar so that they can see what’s happening through their entire forecasting area. But, TDWRs certainly help by covering the cone of silence.

Now that we’ve covered the specifics of the radar network, it’s time to talk about what these radars can actually be used for. But, this article has gone on plenty long as it is. So I’ll be back Monday with a flood of cool images to talk about much more interesting stuff, like hook echoes, and seabreezes, and even bats, in Part 2 of our radar discussion.