NATIONAL WEATHER SERVICE
DOPPLER RADAR
Doppler radar is an exciting new tool forecasters have at their disposal to observe the atmosphere. In this section I will briefly explain what Doppler radar is, how it works, and give some examples of what we can observe with our radar.
HOW DOPPLER RADAR WORKS
All weather radars work on the scientific principle of Scattering. Scattering is the process by which radiation is reflected by small particles and the amount of scattering depends upon the ratio between the particle size and the wavelength of radiation. For example the reason the sky is blue is because air particles (small size compared to raindrops) scatter all colors (small size compared to wavelength of Doppler radar) in the spectrum except for blue. This is called Rayleigh scattering. For weather radar it is the precipitation particles (or other things as we'll see later) that scatter the energy of the radar beam. The wavelength of the Weather Service Radar (also known as WSR-88D) is about 10 centimeters.
The sensitivity of Doppler radar is proportional to the transmitted peak power. A higher transmitted power enables the radar to observe fine structure because more energy will be returned back to the radar. The beam width or cone of energy that is transmitted also determines how much detail the radar can observe. A small beam width enables greater sensitivity than a larger beam width. All National Weather Service (NWS) Doppler radars transmit with a peak power of 750,000 watts and a beam width of 0.88 to 0.96. Older conventional radars had a peak power of 250,000 to 410,000 watts and a beam width of 1.6 to 2.0. Also, NWS Doppler radars have the ability to tilt vertically up to 19.5 in severe weather mode. This enables us to scan mid levels of the storm (15,000-30,000 feet) where the precursors to tornadoes, called mesocyclones, first develop. These features show that the WSR-88D radar system is much more improved over older radars and gives NWS meteorologists a better tool to observe storms. Our expansive radar network covers the entire country with 115 sites. This gives us the ability to dial in to other weather service radars and get radar information on approaching storms from them. These features show that the National Weather Service Doppler radar network is the most versatile and powerful radar network in the world.
Our Doppler radar is based at the Sioux Falls National Weather Service Forecast Office located at the corner of Benson Road and Minnesota Avenue. As you drive by you will see something that looks like a soccer ball perched upon a tower. Located in this ball is a 28 foot diameter dish antenna that rotates in full circle. The antenna is incrementally tilted vertically about 1.0 each time it completes a revolution. The radar system generates a pulse of coherent (fixed frequency) energy that is sent out by the antenna. The energy strikes precipitation particles and some of it is scattered (reflected) back to the radar. The distance of the precipitation echo on the radar display is calculated by the time it takes for one radar pulse to reach the precipitation and return. The strength of the precipitation is measured by the amount of energy (power) returned to the antenna by the reflecting particles.
What's great about our Doppler radar though is that not only is it more sensitive than older radars so we can see more structure in the atmosphere, but it also tells us in what direction, with respect to the radar, precipitation particles are moving. Doppler radar is based upon the fundamental property called the Doppler shift. The Doppler shift can best be described by the analogy of a car moving toward you with the horn blowing. You hear a higher pitch (frequency) as it moves toward you and then a lower frequency as it moves away. The same analogy can be said of the Doppler radar. Precipitation particles that move toward the radar show a different phase shift than those moving away. We can process this data such that the air flow with respect to the radar is displayed as different colors.
The National Weather Service WSR-88D consists of three main components: The Radar Data Acquisition Unit (RDA), the Radar Product Generator (RPG), and the AWIPS workstation. Briefly, the RDA contains the antenna, pedestal, radome, transmitter, receiver, and signal processor. The AWIPS workstation contains graphics and display processors, an archive device for recording imagery, and a printer for printing selected images. In addition to viewing our own radar imagery Sioux Falls meteorologists have access to imagery from many other WSR-88D radars across the central and northern plains so that we can observe weather systems moving into our area.
SCAN STRATEGY - CLEAR AIR VS. PRECIPITATION MODE
All National Weather Service Doppler radars work in either of two modes; clear air mode or precipitation mode. During precipitation mode the radar samples a full volume of the atmosphere every 5 to 6 minutes. That is, the antenna circles a full sweep and rotates vertically from 0.5 to 19.5 every 5 to 6 minutes. To see an example of the scan strategy for the WSR-88D in precipitation mode click here. In clear air mode, the scan strategy is from 0.5 to 4.5 every 10 minutes. So the radar is turning faster and getting more data in precipitation mode than in clear air mode. But the slower rotational rate in clear air mode enables the radar to "sense" the atmosphere in more detail such that we are able to see more structure in both the reflectivity and velocity fields.
During precipitation mode forecasters receive radar information every minute and have a whole suite of products for the entire depth of the atmosphere every 5 to 6 minutes . Not only can we see the strength of thunderstorms and their rotation at different elevations; we can observe how the wind speed and direction is changing with height, the potential for large hail, average reflectivity in layers, forecasted storm tracks, and even an estimate of rainfall amounts in one hour, three hour, and storm total intervals. We can also dissect the storm by comparing horizontal slices of reflectivity and velocity up through the thunderstorm - kind of like a CAT scan of the storm. By doing so, NWS meteorologists can look for certain structures or signatures that indicate physical processes occurring in the thunderstorm which lead to severe weather.
Because of the sensitive nature of clear air mode forecasters can see very weak returns. However, these returns are not necessarily cause by precipitation. While it is true that our Doppler radar can see snow much better than previous conventional radars (snow is a poor reflector) it turns out that we can also see returns from other "particles" floating in the atmosphere. Most of the time it is dust picked up by the wind but, in the summer, it is often bugs and small birds. As weather systems move through, a large concentration of bugs can get caught along a cold front or outflow from a thunderstorm. These "bug scatterers" show up as a line of higher reflectivity that can be tracked along the front as it moves through. During the spring and fall bird migration periods it has been shown by research meteorologists that the movement of birds along the wind can affect the wind calculation algorithms such that the flow displayed is faster than the ambient winds. Forecasters are trained to be aware of this and accurately interpret the data. In the spring, under southerly flow, this will show as a stronger south wind and in the fall, under northerly flow, this will show as a stronger north wind.
FALSE RETURNS - GROUND CLUTTER AND ANOMALOUS PRECIPITATION ECHOES
The radar energy not only interacts with precipitation and clear-air returns, it also hits nearby objects such as buildings and trees which return a large portion of the emitted energy back to the radar since they are so close. This results in an area right around the radar that appears like precipitation bit it doesn't move. This is called ground clutter and the radar is designed to filter out most of this false return.
Another false return that is harder to distinguish from real precipitation is called anomalous propagation. This often occurs when there is a strong inversion (cold air near the surface with warmer air aloft) or from a layer of cool, moist air left behind departing thunderstorms. This change in air density between the layers of air with different temperature and moisture characteristics "bends" or changes the direction the radar beam travels. In the above scenarios the beam bends downward striking the ground leading to ground clutter return. This radar return can be minimized and often wiped out completely by our trained meteorologists who know how to recognize and distinguish this type of false return from real precipitation echoes.
RADAR TRAINING AND HOW SIOUX FALLS METEOROLOGISTS USE DOPPLER RADAR IN THE WARNING PROCESS
Each forecaster who operates the WSR-88D system is highly trained to analyze severe weather using the abundance of severe weather products available on the radar. All forecasters have either spent a month in Norman OK (where Doppler radar was first developed for the National Weather Service), or taken an intensive Distance Learning Course through the NWS Training Center, to learn about the radar system and how to interpret radar imagery. We have been taught by experienced personnel from the university and research community and the military. This enables us to accurately warn the public for impending severe weather - it is our mission to protect life and property. Our forecasters make the decision to warn or not to warn based on our interpretation of Doppler radar imagery. We are responsible for issuing severe weather warnings for a large County Warning Area which includes 45 counties southeast South Dakota, Southwest Minnesota, northwest Iowa, and northeast Nebraska. Once we decide to warn, the warning is sent out electronically to all users. In less than a minute you most likely get the warning after it has been received, processed, and relayed to you by the local media or from one of our NOAA Weather Radio Stations. NOAA Weather Radio is activated by a special alarm that is triggered when the warning is transmitted.
Another feature of our radar is the ability to continuously archive radar data. With this archived information, we can play back severe weather events and meticulously analyze storms to further our understanding of severe weather processes and keep our warning skills sharp. Our forecasters must review past severe weather events before and during the severe weather season, and our management team also uses this data, along with warnings issued, to take stock of our warning services and learn how to improve these services.
EXAMPLES OF WSR-88D IMAGERY
While the radar is most helpful to our staff during severe weather it has many other features that help us forecast daily weather. The case studies presented in the Educational Links and Significant Weather Events section of our homepage highlight some examples of radar structure observed with the Sioux Falls Doppler radar. Below are other examples of imagery we use to help us understand and forecast your weather. For more on Doppler radar and how to interpret radar images, follow this link to the University of Illinois Department of Meteorology on-line tutorial:http://ww2010.atmos.uiuc.edu/(Gh)/guides/rs/rad/home.rxml
REFLECTIVITY - PRECIPITATION MODE
As noted above since the radar operates in two modes, precipitation and clear air, we get reflectivity imagery that will appear different from one mode to the next. This example of our radar in precipitation mode shows severe thunderstorms across northwest Iowa. The color coded bar to the right indicates the strength of these storms measured in dBz. In precipitation mode we can detect returns from +5 to +75 dBz. The higher numbers indicate stronger precipitation echos. Note the thin green lines of weaker reflectivity (around 20 to 25 dBz) surrounding Sioux Falls. These lines of weak reflectivity are a result of old outflow boundaries or differences in temperature and moisture left behind from older storms. In fact you can see more recent outflow boundaries emanating from the storms over northwest Iowa (at arrows) which were moving to the west.
REFLECTIVITY - CLEAR AIR MODE
Under light precipitation events, such as light showers and especially snow, the radar will most likely be operating in clear air mode. This enables us to see more structure in the precipitation patterns because the radar can measure and process very weak returns. This clear air mode example of light snow across our area depicts a swath or band of snow across northwest Iowa with numerous organized bands of snow oriented north to south. These bands were oriented along north winds. The color coded bar to the right shows how strong of a return we are seeing. In clear air mode we see returns from -28 dBz to +28 dBz which is much more sensitive than precipitation mode. It is this sensitivity that allows our forecasters to see fine structure and much more weaker returns than we can in precipitation mode.
VELOCITY
The Doppler radar's strength, its ability to detect wind flow, tells meteorologists much about the atmosphere and the circulation within thunderstorms in particular. Velocity information is a bit more complicated to interpret to the untrained eye but it can tell skilled meteorologists a wealth of information. The key point to remember is to always interpret the velocity information with respect to where the radar is located on the image. For example, this velocity image from the same clear air mode return as above shows us what the snow bands look like from a velocity perspective. The key to interpreting this image is the color coded bar to the right. Green and blue colors (negative numbers) show air moving toward the radar while reds and yellows (positive numbers) show flow moving away from the radar. In this example there were north winds across the entire area. The radar data indeed shows greens to the north of the radar (inbound) and yellows to the south of the radar (outbound). The strength of the winds are measured in knots. In this example the strongest inbound winds are in dark green; about 26 to 35 knots while the strongest outbound winds are in red; about 36 to 49 knots.
In this example the scatterers were more widespread (most likely bugs given the time of year; August) giving us a more complete picture of the wind flow. Here the wind is from the northwest because blues (inbound) are northwest of the radar and yellows (outbound) are southeast of the radar.
We can loop these velocity images and see how the wind is changing with respect to the radar. This comes in handy, for example, when observing a cold front moving through the area. In this example a cold front is moving from northwest to southeast over the radar site. The inbound winds (green and blue) and the outbound winds (red) associated with the wind flow ahead of and behind the front also change as the front moves over the radar.
This example shows us how we would detect wind flow in a thunderstorm. For a radar located to the south the top velocity image would indicate cold air spreading out from beneath a thunderstorm (what we call divergence). The bottom image shows cyclonic rotation (counterclockwise rotation) which would indicate a rotation in the thunderstorm that could be the precursor to tornado formation.
WIND PROFILE DATA
The wind information can be displayed at any elevation angle letting us know the wind speed and direction at different heights in the atmosphere. One product that lets us see this data more easily is the wind profile. We can display a profile of the wind above the radar site for each volume scan; the depth of the atmosphere the radar senses. In this example we are seeing a time series of wind speed and direction at different heights in the atmosphere. Height is in thousands of feet and time increases from left to right with intervals every 6 minutes. This example shows a nice frontal inversion characterized by a wind shift about 3,000- 4,000 feet. You can see a shallow layer of strong northeasterly flow at 3,000 feet and below with a deeper layer of southwest flow above 4,000 feet. This rapidly changing wind direction with height is an indication of wind shear that is important to airport operations and pilots.
FOUR PANEL DISPLAYS
Warning forecasters can display radar information in a number of ways to help them interpret imagery and look for certain structures that occur with severe weather. Looking at one low level slice of the storm will not tell you that much. A skilled radar interpreter will look at different slices of the storm vertically and compare them to see how the reflectivity and velocity structures change with height. One way to do that is with a 4 panel display and a cursor that is linked geographically to the same location in all 4 panels. This is analogous to a CAT scan because you are seeing horizontal slices of the storm at different levels. This example shows a 4 panel reflectivity image of a severe thunderstorm near Beresford. Quadrant 1 (upper left) is at low levels of the storm with quadrants 2 and3 successively above. Quadrant 4 (lower right) is around 27, 000 feet. You can see how the reflectivity structure changes with height. The low levels show a classic hook echo (like the number 6 on its side) defined by high reflectivities (red colors) that curl around a weak reflectivity notch (blue and green in quad 2). In the mid levels of the storm (quad 3) we see a hole of weak reflectivity surrounded by much higher reflectivity. We call this feature a Bounded Weak Echo Region (BWER) which indicates physical processes in the that are causing organization and a rotating updraft.
This 4 panel display is of another severe storm that dropped 4.5" diameter (softball size) hail near Salem. A particular feature of this storm that indicates its severity is a pronounced spike of weak reflectivity jutting out (shown by arrow) from the core of the storm in quadrant 3. This is called a flare echo and is caused by the reflection of very large hail in the mid levels of the storm. You can also see a BWER in quadrant 4 over a reflectivity notch in quadrant 3 (just beneath the "A" in Canova indicating a rotating updraft with counterclockwise circulation.
The velocity data for this storm a short time later in 4 panel display shows us how the wind flow within this storm is changing with height. We can observe and measure winds flowing in opposite direction that indicates rotation in the storm. In this example, quadrant 3 shows blue and yellow colors right next to each other indicating inbound and outbound winds in juxtaposition. This is very strong horizontal wind shear indicative of a strongly rotating column of air we call a Mesocyclone. In fact the yellow circle is where the radar algorithm also has identified a mesocyclone within this storm. Forecasters can measure this speed differential (called rotational shear) to gauge the circulation's strength. A strong mesocyclone could lead to the formation of a tornado but in this case it did not.
COMPOSITE REFLECTIVITY AND STORM TRACKS
The Composite Reflectivity gives us the maximum reflectivity measured throughout the depth of the storm. It also has a table at the top showing us various attributes of individual storms such as mesocyclone and tornadic potential, probability of severe hail and size, the height of the storm, and its forecast movement. The movement is indicated by a thin black line in this example and shows most storms moving from southwest to northeast. However there are two storms just to the north of FSD that are moving to the east and southeast. Storms that move in different directions from other storms indicate physical processes in the storm that produce a rotating updraft. This can lead to severe weather.
CELL TRENDS
The cell trends chart is a new addition to the WSR-88D system and is used in conjunction with the composite reflectivity chart to give us a trend of different attributes for a particular storm. It is a graph of the storm attribute vs. time and shows us whether its intensity is increasing or decreasing with time. In this example the cell trends shows all attributes of the storm increasing with time. The storm top is growing to 50,000 feet (quad 1), the probability of severe hail (>3/4" diameter) is rapidly increasing to 100%, the cell based VIL (a measure of the liquid equivalent in the storm) is gradually increasing, and the maximum reflectivity is slowly increasing. All of these indications taken with observed reflectivity and velocity structures in the storm help us decide whether to warn (as in this case) or not.
ECHO TOPS
The echo tops product graphically shows us the height of thunderstorm tops in thousands of feet. In this example there is an isolated thunderstorm embedded in an area of low topped showers.The color bar to the right shows us the storm top is 55,000 feet high.
VERTICALLY INTEGRATED LIQUID (VIL)
This product is an estimate of the liquid water content in a storm based on an integration of reflectivity in a vertical column through the storm. Higher VIL readings indicate an increasingly higher potential for severe weather and flooding. This example shows a time lapse in the four panel format with time increasing from panel 1 to panel 4. Red colors indicate very high VIL values and a storm with a large water content. Very heavy rain and a downburst of strong winds accompanied this storm as it unleashed its fury near the Sioux City area.
PRECIPITATION AMOUNTS
A useful tool to help forecasters warn for flash flood events is the precipitation amounts graphic. These amounts are estimated by the radar using a unique equation that relates radar reflectivity to rainfall rates. These amounts can be displayed in one hour or three hour increments, and as a total amount for the storm's duration. This example shows a heavy rainfall event in June, 1996 where there were widespread rainfall amounts of an inch or more and portions of northwest Iowa received around 6 inches of rain.
Original document written by Ron Holmes and posted
January 9, 1998
Slight Modifications to this page were made on: October 7, 2004
