|
TECHNICAL EXPLANATION OF NEXRAD
If you are interested in a brief explanation of NEXRAD, please
click here. The explanation that follows is
significantly more in-depth.
QUICK OVERVIEW
Courtesy of NOAA
|
As covered in the brief explanation and in its
simplest form, NEXRAD works by sending out a pulse of energy (the green wave in the
animation to the right) and detecting the "echo"
or "return" (the blue wave) that is received from airborne objects such as rain, hail,
dust, smoke, snow, etc.
The time it takes for the echo to return allows the radar to determine the distance to
the object while the minor difference in the wavelength (caused by the Doppler effect)
allows it to determine the speed of the object relative to the NEXRAD station.
That's the quick explanation. The rest of this page will explain the process in more
detail.
THE RADAR DOME
NOAA
|
The Radar Dome (left) is what most people think of when they think of NEXRAD. The dome is over
30 feet in diameter and usually on a tower over 100 feet high so they can often be
seen from quite a distance and can look somewhat ominous and mysterious when seen up-close.
It is made of a material that allows the radar's signal to pass through relatively
unchanged.
The dome itself mostly contains air and is actually just a protective housing for the
much more familiar-looking 28-foot radar dish (below) which it encloses. Within this
protective housing, the radar dish constantly rotates just as all other radars we
have seen on top of airport control towers, navy ships, and in movies.
So, while NEXRAD sites may look very different from other radar antennas that we have
seen, it's really the same idea. They just happen to have comparatively big dishes
enclosed in a big white ball on the top of a tall tower.
POWER SUPPLY
A NEXRAD site consumes approximately 50.8 kilowatts of energy. That includes the air
conditioners and/or heaters to control the temperature of the hardware. The
transmitter itself takes 15 kilowatts--roughly the same amount as 13 clothes irons.
NEXRAD radar draws its power from the normal power grid. However, it has a very robust UPS
system to keep the radar operating even in the event of a loss of power. Imagine if the
radar stopped operating because of the storm it was monitoring knocking out its power!
THE TRANSMITTER (KLYSTRON)
Klystron, NOAA
|
Although a NEXRAD site consumes about 50.8 kilowatts of power, the klystron (pictured to the
left) is a special (and expensive) unit responsible for converting standard commercial power
to the 750,000 watts of coherent energy that is transmitted in each pulse. It consumes
7.5 watts, gets hit with 47,000 volts, and amplifies it into the 750,000 watts needed for each
pulse. The klystron takes the place of the magnetron used in older radar systems.
The peak effective power of a NEXRAD site is, as mentioned above, 750 kilowatts. However,
the transmitter is only active between 0.05% and 6.1% of the time. As mentioned in the
brief introduction the radar is only transmitting for about
7 seconds out of each hour--or 0.19% of the time. As such, the average output power of a
NEXRAD site is 0.0019 * 750,000 = 1458 watts. The specifications put the average output
power is 1300 watts. In reality, the exact average output power depends on the mode in
which the radar is operating.
THE RADAR DISH
Dish, NOAA
|
Within the radar dome,
the 28-foot radar dish constantly rotates clockwise to scan the sky in all directions around the
NEXRAD site. The rotation of the dish is powered by a very small DC stepper motor about
1-1/2 feet long. The radar dish itself is so well balanced that the small
motor can power the rotation--in fact, it is easy to rotate the dish with just one hand!
Unlike traditional radar, the dish doesn't scan the same position over and over but rather adjusts
the angle of the dish after every rotation. For example, the lowest angle is 0.50 degrees
above the horizon. This is low to the earth and allows the radar to detect weather far
off in the distance. This same angle, however, will not give the radar an image of
weather that is much closer since it will literally undershoot it. For this reason the
radar can scan up to an angle of 19.5 degrees.
Thus the radar dish is constantly rotating and also adjusting it's vertical angle over
the horizon.
RADAR BEAM STRUCTURE
The beam emitted by the radar is generally conical in shape, but is not a perfectly
focused line of energy as if it were a laser. It is a beam of energy that is 0.96 degrees
in width as pictued in the diagram below.
Radar Beam Structure, NOAA
|
The width of a radar beam is measured by the angle between the "1/2 power points." The main beam is at
full power in the middle of the beam as it leaves the dish. At 0.48 degrees in
each direction the energy is one-half as strong as it is at the
center of the main beam. Thus, the beam of energy emitted by the radar is
wider as its distance from the radar dish increases.
| Distance |
Beam Width |
| 50 miles |
5041'
|
| 100 miles |
10,082'
|
| 150 miles |
15,124'
|
| 200 miles |
20,165'
|
| 250 miles |
25,206'
|
The width of the beam is directly proportional to the distance from the dish based on
the calculation Width (in feet) = Distance (in miles) X 100.82. Essentially,
the width of the beam increases by 100.82 feet for each mile the distance to the radar
increases. Thus at 5 miles, the beam width would be 5 x 100.82 = 504.1 feet. The
table to the right shows the beam width at 50 mile intervals.
Of course, not all the energy is contained within the focused beam. Outside the beam,
as defined by the "1/2 power points," there is still some amount of energy, but it quickly
tapers off. Sidelobe energy is energy emitted directly from the dish in directions
other than that of the intended beam. This results in false returns from nearby targets,
though the returns are so weak that they generally do not affect the performance of the
radar
Twice each year (once in spring with the new leaves and once in
fall after the leaves fall) the technicians run special diagnostic software to identify
the recurring ground clutter. This allows the system to automatically filter a great deal
of the ground clutter that would otherwise seriously pollute the image near the radar
site.
PULSE LENGTH & REPETITION TIME
The beam itself is not a cosntant stream of energy but a series of "pulses." Each pulse
consists of a short period of time, called the "pulse duration," during which the
transmitter is active. The radar then stops transmitting and waits a certain amount
of time to listen for an echo. The total amount of time from the start of one pulse
to the start of the next pulse (including the "listening time" in between) is the
"pulse repetition time."
Pulse Diagram, NOAA
|
In the diagram to the right, the pulse duration is 1.57 microseconds (0.00000157 seconds).
Since the speed of light is 300,000km per second we can compute the wavelength by multiplying
300,000km by 0.00000157 seconds for a calculated length of approximately 500 meters
(471 meters to be exact).
The radar then stops transmitting and listens for an echo during the "listening period"
which, in the diagram to the right, lasts for 998.43 microseconds. The pulse duration
of 1.57 microseconds plus the listening period of 998.43 microseconds together form the
pulse reptition time of 1000 microseconds which is the same as 1 millisecond, or 1/1000th
of a second. Since the radar, in this example, is sending out a pulse every 1/1000th of
a second, it is sending out a total of 1000 pulses per second.
NEXRAD transmits in either "short pulse" or "long pulse" mode. In short pulse mode, the
pulse duration is 1.57 microseconds as in the previous example. When in long pulse mode,
the pulse duration is extended to 4.7 microseconds for a wavelength of approximately
1500 meters (1410 meters to be exact). The longer the pulse, the more energy will be
scattered by any given target and the more energy will be received back by the radar. Thus,
NEXRAD is more sensitive when operating in long pulse mode.
MAXIMUM RANGE
The maximum range of the radar is directly proportional to the pulse repetition time. The
longer the time between pulses, the greater the range. This is due to the fact that the radar
beam is traveling away from the radar at approximately the speed of light. The radar
creates its scan by timing how long it takes the radar beam to hit an object and echo
back. The longer it listens between pulses, the more time is available for the beam to
travel away from the radar and bounce back.
In the example above, where we had a pulse repitition time of 1 millisecond (1/1000th of a
second), we may calculate how far the beam can travel in that time by multiplying 1 millisecond
(0.001 seconds) by the speed of light (300,000km/second) for a result of 300km.
However, keep in mind that the beam has to be able to reach its target and reflect back
in that time which means the total round trip distance is 300km. That means, with
a 1 millisecond pulse reptition time, the total range is half that: 150km.
NEXRAD uses pulse repetition frequencies of between 318 and 1304 pulses per second, which
translates to pulse reptition times of between 3.144 milliseconds and .766 milliseconds
(766 microseconds). Therefore, the range of NEXRAD is between 471.6km and 114.9km.
DETERMINING THE DISTANCE TO OBJECTS
Determining the distance to an object (raindrop, hailstone, dust, etc.) works the same
way as computing the maximum range. The radar times how long it takes for an echo to
return, multiplies by the speed of light, divides by 2 (since the beam had to reach the
object and come back), and that's the distance.
In the graphic example to the left, the radar's beam bounces off a raindrop within the
cloud and is detected by the radar 425 microseconds (0.000425 seconds) after it was
sent. By multiplying the measured time by the speed of light we know that the beam
covered 127.5km and we know that half of that distance was the distance to the
cloud and the other half was the distance back. So we know the raindrop we
detected is 63.8km away.
This process is repeated for each and every "echo" that is returned to the radar. This
is a truly staggering amount of data when you consider that each and every bird, insect,
hailstone, raindrop, dust particle, or pollen can potentially reflect some amount of
the energy back to the radar.
RANGE FOLDING
Range folding is an interesting error that can occur when the radar detects something
(raindrop, snow, hail, etc.) beyond its maximum range and is erroneously reported as
being much closer.
In the example above in the "Maximum Range" section, we had a pulse repitition time of
1 millisecond (1/1000th of a second) which produced a maximum range of 150km. The maximum
range was calculated as such since after 1 millisecond the radar will stop listening and
transmit the next pulse. 1 millisecond was sufficient time for the energy to complete a
300km round-trip, half of which is the maximum range.
But what would happen if the radar transmits a pulse and the energy were to hit an object
180km away? That would require a 360km round-trip which, at the speed of light, would take
1.2 milliseconds (600 microseconds in each direction). So the radar would send out a pulse and
600 microseconds later the pulse would hit the object at 180km and reflect back. However,
400 microseconds later (a total of 1 millisecond after the first pulse was transmitted) the
radar would send out its next pulse. 200 microseconds later the reflected energy from
the object 180km away would be received by the radar. However, the radar would believe that
the energy had taken only 200 microseconds to return since that is the time that had
elapsed since the last pulse was sent. A 200 microsecond round-trip equates to 60km, half of
which is 30km. So the radar would erroneously calculate the distance to the object to be 30km
instead of 180km.
This phenomenon is known as "range folding" and causes real objects in the atmosphere to appear
to be located much closer than they are.
DETERMINING THE VELOCITY OF OBJECTS
In addition to its high sensitivity, one of the main features that sets NEXRAD apart from
previous weather radars is the fact that NEXRAD is a Doppler radar. This means it's
not just capable of detecting an object in the atmosphere, but also its speed relative to
the radar site.
Just as the Doppler Effect causes the apparent tone of a horn to
change as a car drives past you by changing the wavelength of the sound wave, the movement
of an object in the atmosphere will cause a slight change in the wavelength of the
radar's return energy. NEXRAD can then detect this minor change in wavelength and calculate
the relative speed of the object. This is the exact same principle that allows a police radar
to determine your speed as you drive by (well explained here).
Inasmuch as the Doppler effect is concerned, the return energy of the radar can be considered to be
emitted by the object in the atmosphere. So just as a car driving towards you or away from
you honking his horn will have a different sound wavelength, the energy returned to the radar from
the object in the atmosphere will have a different wavelength, too.
The wavelength of the echo will differ, or shift, based on the equation Frequency Shift =
Object Velocity X Radar Frequency / Speed of Light.
So if we have a raindrop being blown in the wind towards the radar at a speed of 30 miles/hour (48 km/hour =
48,000 meters/hour = 13.3 meters/second),
we multiply that speed (13.3) by the radar's frequency (3GZ) and divide by the speed of light
(300,000 km/second) to arrive at a value of 133. So while the original frequency sent by the
radar was 3,000,000,000 Hz, the frequency of the echo returned from the raindrop is
3,000,000,133 Hz. Since the raindrop is moving towards the radar the waves of the echo will be slightly
compressed, resulting in a higher frequency. If the rain was being blown away from the radar site,
a speed of -13.3 would be used instead and the frequency of the echo returned from the raindrop
would be 2,999,999,877 Hz.
- If the object is moving towards the radar, the echo frequency will increase.
- If the object is moving away from the radar, the echo frequency will decrease.
- If the object is moving parallel to the radar, the echo frequency will not change
It's very important to note that the speed detected by the radar is the speed relative to the
radar. If an object is heading directly for the radar (and will eventually pass over it) then
the detected speed will be the actual speed. However, if the object isn't heading directly at the radar
then the speed detected will not be its actual speed but rather its relative speed.
BEAM ANGLES & DISH ROTATION
A complete NEXRAD image (called a "volume scan") is made up of multiple scans (called
slices) at multiple angles. The exact number of slices and at what angles depends on the
mode of operation under which the radar is currently operating. Each slice requires
one rotation of the dish. Many slices combined form a volume scan which is
what is eventually presented to the end-user.
VCP21 "Precipitation Mode", NOAA
|
In the "Precipitation Mode" (shown to the left)
--also known as VCP21 mode--the radar combines a total of 11 slices at 9 separate
angles (the 2 lowest angles are scanned twice) into a single volume scan. This places
emphasis on quickly establishing local conditions. The dish rotates relatively quickly,
completing a volume scan in a total of 6 minutes with the radar completing a complete
rotation approximately once every 33 seconds. Since this mode is used primarily when
there is nearby precipitation, the sensitivity of the radar is reduced. That is to
say, relatively faint signals are generally not visible in this mode so that full
attention and importance may be placed on the already-developed weather system moving
through the area.
VCP31 "Clear Air Mode", NOAA
|
In "Clear Air Mode" (shown to the
right)--also known as VCP31 mode--the volume scan consists of a total of 7 slices at 5
separate angles (the lowest 2 are scanned twice). The speed of the radar's rotation is
significantly reduced to about one rotation every 86 seconds, creating a complete volume
scan in about 10 minutes. Since there is no significant meteorlogical activity when the
radar is in this mode the radar doesn't bother to do scans at the highest angles and
instead focuses carefully on long-range, more sensitive scans to detect forming weather
patterns. The slower rotation of the dish gives the radar increase sensitivity in this
mode.
These two diagrams tell you at what elevation the radar beam will be at a given distance
from the radar site. For example, the 0.50 beam will be scanning at an elevation of
about 10,000 feet when it is 120 miles from the site while the 4.5 beam will be scanning
at an elevation of approximately 70,000 feet at the same distance.
Visual Depication of Beam Angles
|
Of course, the angles aren't as severe as they appear in these two diagrams since the
vertical altitude and the horizontal distance aren't to scale. The diagram to the left gives
you a better visual idea of the approximate angles of each
The dish itself never stops rotating. The speed of rotation will vary depending
on the mode in which it is operating, but it never comes to a complete stop as this
would cause additional strain on the moving parts and the electric motor, not to mention
that a radar not in use is a useless radar. :)
Sources and Acknowledgements:
| 
