Radar, Lies and Videotape, Part. II

Yesterday we talked about how radar maps can lie about snow falling (and snow not falling). During review of the "raw" regional radar loop that I posted, I noted that there were some other interesting non-weather radar artifacts to be revealed. Because AccuWeather typically post-processes their radar composites, you wouldn't normally see these on AccuWeather.com but we make the raw composites available for radar and meteorological researchers.

At 05:55Z time, a bright burst of what might appear to be a large area of thunderstorms, blew up over southeast North Carolina.

If you look at the local radar which produced the phenomenon, it's more clear that this is not, in fact, thunderstorms. The "storms" are very granular, and there are many instances where a very high dBZ value is located next to nothing; normally storms will gradually increase through the dBZ color scale before reaching high values.

Note also that the echoes are all over land - they all stop at the water. The biggest clue though, is if you look at a loop -- the area increases and decreases wildly, behaving unlike slower severe storms. Another clue is that the radar was in "clear-air mode" when the phenomenon started(you can see the words "Clear-Air Mode" in the upper left -- this mode also makes the colors seem brighter; note in the keys however, that the values are the same between clear-air and regular "precip" mode).The NWS [JessePedia] normally only puts radars in this mode if precipitation is not expected.

What's happening here is that the radar beam is bouncing off a stable layer of the atmosphere and hitting the ground, reading elevation, trees and buildings. That's why the "echoes" stop at the shore.

At the end of the loop (11:55 Z time), you see a series of "spikes", appearing first in the east, then travelling west. This is the sunrise. When the sun first comes up over the horizon, it temporarily freaks out the NEXRAD radar system, making it think that there is a narrow band of energy. Here's what it looks like on the local radar.

Spikes like this happen randomly when the weather is clear, usually indicating that the beam is bouncing off something nearby (one radar in Florida was rumored to have been located too close to the metal roof of a large barn, causing a persistent spike). Here's the local radar for the Pittsburgh spike you see below, which happened at 2255 Z.

You can also see a lot of "blobs" in the Midwest in this image. Notice that they are lined up with the individual NEXRAD Doppler radar locations. This extremely common occurrence happens when each radar's beam detects objects such as trees, buildings and mountains on its way out from the radar transmitter. Here's an example of what it looks like up close on the Indianapolis radar.