Hello:

Here is a Radar FAQ from MIT. For the more technically oriented perhaps this will spark some theories on the ring anomalies.

http://www-paoc.mit.edu/Radar_Lab/FAQ.html#6

Frequently Asked Questions

How do radars work? Radars work by "shining" a beam of energy (usually microwaves) at an object and measuring how much power gets reflected back. The object might be an airplane, precipitation in a thunderstorm, bugs in the atmosphere, etc. If the object is moving when the energy hits it, then the microwaves will be Doppler shifted after they have been reflected back, and we can partly measure the object's velocity.

How much energy do radars transmit? Is it dangerous? This varies with the radar, but the answer to the second question is almost always no. Weather radars might transmit from 100-500 kW peak power, but they only do this in very short bursts. For example, the MIT C-Band radar transmits 150 kW peak power, but only in 1 microsecond pulses, 1000 of these per second. So the average power over any 1 second interval is only (150,000 * 10^-6 * 1000) kW or about 150 Watts - perhaps as strong as a couple of bright light bulbs. Further, the intensity of power in a microwave beam drops off as 1/(range^2); by the time a weather radar beam hits a target and bounces back, the antenna usually measures at most milliwatts of power.

Nonetheless, OSHA regulations on safe levels of microwave exposure are somewhat sketchy. Radar engineers working close to the antenna dish typically prefer the transmitter to be off, but more than a few hundred meters away, the intensity has almost certainly dropped to safe levels.

What is it that weather radars "see"? By the time a weather radar beam hits a storm, it is typically several hundred meters wide. It thus reflects off a volume of cloud which contains many condensed droplets, drops, snow, ice and hail. If the radar wavelength is large compared to the size of this 'condensate', the reflected power will be a useful function of the number and sizes of these particles. Weather radars thus give us a "bulk" picture of a storm: they can tell us how much 'stuff' is in the cloud, but can't directly tell us what size or type (rain, snow, ice, hail) it is. However, decades of research in radar meteorology have taught us how to interpret this "bulk" picture and relate the reflected power (also known as reflectivity) to rainfall rates, hail, the probability of lightning, microbursts, and more.

What do all those colors mean on the TV radar images? Those colors are a way of plotting up the different values of reflectivity (see #3, above) in a storm. The National Weather Service has established a set of five standard 'Levels' of reflectivity, to correspond to the types of rainfall usually associated with a given reflectivity - e.g., drizzle, light rain, heavy rain, etc. Most radars can actually see much more detailed distributions of reflectivity; the NWS 5-Level Scheme was constructed to simplify widespread distribution of radar data.

What is NEXRAD? NEXRAD (Next Generation Radar) is a generic term for the National Weather Service's program of implementing about 160 WSR-88D radars around the country. The current network of NWS weather radars is woefully outdated. After several years of contractual and technical difficulties, deployment of the WSR-88D's is finally underway in force. New sites are coming online at the rate of several per month, and full deployment is expected by 1995. An integral part of the WSR-88D program is the development of a data processing and distribution network. In addition to the raw reflectivity and velocity data which the radars collect, the sites will generate a suite of analyzed products, including:

Combined shear, severe weather analyses, echo top heights, weak echo regions, accumulated precipitation (1,3 hours, storm total), wind retrievals, vertically integrated liquid, storm tracking data, and more.

What's all this (L,S,C,X,K)-Band nonsense? The letters refer to archaic military designations of the different frequency or wavelength 'bands' that these radars operate in. They harken back to the early development of radar in WWII and somehow have stuck with us. They correspond to the following: L-Band: 1-2 GHz, 15-30 cm wavelength. Mostly used for clear-air turbulence studies. S-Band: 2-4 GHz, 08-15 cm wavelength. Used for long- and short- range weather surveillance. The WSR-88D's are S-band. Not easily attenuated but require large dishes and motors. C-Band: 4-8 GHz, 04-08 cm wavelength. Used for short-range weather surveillance (e.g., near airports). Portability means they're often used in research field programs. Nice tradeoff between X- and S-Bands. Nearby bands are often used for microwave communications links. X-Band: 8-12 GHz, 2.5-4 cm wavelength. Used for very short-range work; very sensitive to smaller particles and thus useful for studies of early cloud development. However, attenuated rapidly as they pass through storms. Share some space with police speed radar. K-Band: 12-18, 27-40 GHz, 1.7-2.5, .75-1.2 cm wavelength. Actually two bands, split down the middle by a strong water vapor absorption line. Similar comments as with X-bands, above. Also share space with police radar. What is ground clutter? "Ground clutter" is a term used to describe the (usually large) amounts of microwave energy reflected back to the radar from stationary objects on the ground like towers, hills, high tension lines, trees, buildings, etc. You might ask: if radars are mounted on towers and look out horizontally, why should they see these objects at all? Shouldn't the earth's curvature get rid of them? The answer is yes and no. Although we may point the radar beam horizontally above the tallest obstacles nearby, it is a beam and widens with distance, sometimes intersecting the ground. Also, it is not a perfect "pencil beam"; antenna side lobes weakly transmit and receive power outside of the main beam, and since the ground is very reflective, it may be seen through these side lobes. Finally, local temperature conditions may set up conditions which bend the beam slightly, forcing it to intersect the ground. In Boston we frequently see this at night when our beam passes over the Boston Harbor, is bent downwards, and hits the [relatively flat] Cape Cod peninsula. We can often trace out the Cape exactly from the ground clutter! (Click here to see)

Fortunately, ground clutter is often removable. Since it is usually stationary, we can recognize it as highly reflective but unmoving echoes, and filter out or attenuate the signals. However, in high-relief environments, we don't want to blank out large regions because of ground clutter. Clutter removal techniques, both in hardware and software, are one of the more active areas of research in radar technology.

How well can radars quantitatively measure rainfall? Rainfall is typically estimated from radar data with a Z-R relationship, which converts bulk reflectivity to a rainfall rate. Several "stock" Z-R relationships exist, for different types (convective, stratiform, winter, tropical) of storms. Direct use of the "stock" relationships may give errors of O(50-100%). Use of more sophisticated correction techniques, or local measurement of precipitation drop size spectra, may bring these down to O(30-50%). Given that rain gauges can typically measure rainfall to O(5-10%) accuracy, this may seem pretty bad. However, the real advantage of radar is that it can measure the full extent of the precipitation field over large areas, whereas gauge networks are often widely-spaced point measurements. For hydrological applications such as basin modelling and flood prediction, this makes radar data competitive with rain gauges, particularly when integrating over the life of a storm.

How well can Doppler radars measure winds? Winds form a fluid flow in three dimensions - say, E/W, N/S, and up/down. They thus have 3 components which need to be measured. However, radars measure a Doppler shift in only one direction - towards or from the radar, along whatever path the antenna happens to be pointing. As such, a radar offers an incomplete picture of the actual wind field. How do we get around this? Well, near the surface, the radar looks essentially horizontally, and the component of the Doppler shift from vertical motion (falling rain or microbursts) contributes only a tiny bit to the overall Doppler shift. So we're down to 1 measurement for 2 dimensions.

If we make some assumptions, we can further estimate the general wind field around the radar, but have difficulty getting all the information about local scale circulations (gust fronts, outflows, etc.). The VVP and VAD techniques make use of this approach.

Fortunately, these so-called radial winds are still useful, and many effective hazardous wind detection schemes have been developed using radial winds and reflectivity alone.

A further possibility is to use more than one Doppler radar: if we put another radar at a second site, we can (with a little math) extract the second wind component. With a little physics, we can then back out the missing third wind component. Or, we can add a third radar, and actually measure the third wind component. We can thus have single Doppler, dual Doppler, or triple Doppler measurement of winds. Click here to read about such a "triple-Doppler" experiment in Orlando, Florida.

Needless to say, putting three Doppler radars in close proximity to each other is a very expensive way of getting three-dimensional wind fields. The BINET project currently underway at the National Center for Atmospheric Research is exploring ways to get 3-D wind fields from a single transmitting radar and multiple cheap, passive receivers.