Reading Composite Reflectivity and the Storm Attribute Table

This data table corresponds to the radar graphic above. Click on a column to read that help topic.

Basically, composite reflectivity plots the highest dBZ value within a column while the lower values are ignored. Therefore, it may look like it is raining on the ground heavier than what is actually occurring. This product is most useful when the radar is in storm mode (VCP 11), which updates every five minutes, or precipitation mode (VCP 21), which updates every six minutes. Only the radar operator can select the mode. VCP is the abbreviated form of Volume Coverage Pattern.

The purpose of this help file is to allow those new to the composite reflectivity imagery and the storm attribute table learn how to use it effectively. It has been put together using credible web sites on radar interpretation throughout the Internet. Remember that all information that is retrieved from radar algorithms should be used as a guide, as it is not the exact truth. The link between a weather radar and a thunderstorm is full of obstacles. This opens a margin of error. Nonetheless, output from a radar should not be taken lightly either. The several strengths and weaknesses of using composite reflectivity imagery and the storm attribute table follow.

Cell Identification – The National Weather Service radar assigns each thunderstorm that develops or moves into the range of a particular site, within the composite reflectivity product, a tracking code. This code consists of one letter followed by one number, such as in X1. Use this code to locate the cell on the radar. StormLab© only labels the fifteen strongest cells on the radar image itself. Others can be found on the storm attribute table.

Sometimes a cell may not be visible within the 124 nautical mile (124 NM) sweep, but still appear on the corresponding storm attribute table. If this occurs, either switch to the 248 nautical mile (248 NM) sweep from the same site, or select a different site which would show the thunderstorm cell more effectively. If a different radar site is selected, the cell identification code will likely change, as well as the characteristics of the storm.

The reason some of the characteristics may be different pertains to the vertical level at which the storm was scanned, which depends on the cell's distance from the radar. Best imagery is usually taken close to a radar site, but not above it, allowing for all of the beams sent by the radar to reflect off the cell.

Any given thunderstorm may or may not retain the cell identification code between radar scans. This is largely dependent on how consistent the storm is, by means of development and growth, between runs. If the radar believes a new cell has formed, it will identify it with a new tracking code, even though it may have been a detected storm in previous runs that changed structure. Two tracking codes could be assigned to one storm if it appears to the radar that the cell is splitting into two new storms. A linear squall line or bow/comma echo usually has several identification codes.

A thunderstorm that is very near or above the radar site may not be detected because the radar's beams are unable to tilt at such an angle. A cell identification code will likely not be assigned to such a cell, which will tend to have poor and inaccurate reflectivity. This is due to the "cone of silence" effect. Even cells that had a tracking code at one point may lose it.

Note: One nautical mile is equal to 6076.115 feet, or 1852 meters, in the United States.

This storm attribute helps to locate a certain cell on the radar's composite reflectivity product.

Az, which is short for azimuth, tells the direction the cell is from the radar in degrees. For example, an az, or azimuth, of 180 degrees means that a particular storm is directly south of the radar site.

Ran tells the distance that a cell is from the radar in nautical miles. For example, on a 124 nautical mile sweep, a cell with a ran of 62 would appear exactly between the radar site and the perimeter of the radar sweep. StormLab's© unique data sampling tool allows users to identify the exact amount of nautical miles any given location is from the radar site.

The "Az/Ran" of each detected storm is plotted on the radar image, if desired by the radar operator, with the cell identification code just beneath it, and the projected path of the storm ahead of it.

The "cone of silence" may limit the ability of a radar to plot a thunderstorm cell nearby to the transmitter. This occurs because the radar's beams are unable to penetrate the updraft of a storm at such an angle. Therefore, the radar is unable to detect the storm and plot the centeroid of it on the composite reflectivity.

Note: One nautical mile is equal to 6076.115 feet, or 1852 meters, in the United States.

The abbreviation TVS stands for Tornado Vortex Signature. One of three terms will be used to describe the tornadic potential of a certain cell in this column.

Remember that a radar cannot scan the ground to look for a tornadic rotation, which can lead to detection errors. The radar only obtains data from the storm tower, and relies heavily on this to detect a tornado vortex signature.

Since the algorithm which detects the tornadic, or tornado, vortex signature is sensitive, false signatures can at times be detected along frontal boundaries and squall lines, or where a definite wind shift is present, especially when the front edge of a squall line is parallel to the radar beam. A TVS or ETVS does not guarantee a tornado.

When a TVS or ETVS is detected, the radar has found an area of intense and concentrated rotation, more so than a mesocyclone. Watch this for persistence. Other algorithm criteria, such as the strength and vertical depth, must also be met for a radar to indicate a tornado vortex signature.

A TVS or ETVS itself is not a visually observable feature. The possible tornado that could accompany such a signature obviously is visible however.

A radar cannot detect a tornado vortex signature well, or possibly at all, if a thunderstorm is close to the radar site. Beams are unable to extend into the updraft at such a degree from the radar's transmitter. A lesser, or possibly absent, tornado vortex signature is displayed as a result. This creates a feature known as the "cone of silence" on the radar screen.

With that said, it is also possible for storms to weaken. Therefore, it is a good practice to check the atmospheric conditions before judging that the "cone of silence" is responsible for the weakening or dissipated tornado vortex signature.

Do not question warnings issued by the expert National Weather Service meteorologists. If you feel your life is threatened by a thunderstorm, do not postpone seeking shelter, even if a warning is not in effect for your area.

The term MESO is short for mesocyclone, which is generally a radar term. A mesocyclone is a rotation found within the tower of a strong thunderstorm, and can sometimes lead to the formation of a tornado, or a tornado vortex signature on radar.

When a radar detects a mesocyclone, it is good practice to watch for time continuity between volume scans, a diameter of two to six miles, and an extension vertically of over 10,000 feet. These characteristics can only be found in velocity imagery. Nonetheless, the mesocyclone algorithm is fairly good at making sure all of the needed criteria are roughly met. One thing that can be watched from composite reflectivity is the time period at which a mesocyclone is present. The longer the period at which a rotation is sustained within the head of a storm, the better chance there is a mesocyclone is active within that given cell.

One of four key words can be found within the MESO column of the storm attribute table.

Radars can also sometimes falsely detect a mesocyclone if a thunderstorm is directly over a radar site. On the flip side of that, a radar cannot detect mesocyclones well, or possibly at all, if a thunderstorm is close to the radar's transmitter. Beams are unable to extend into the updraft at such a degree from the radar site. A lesser, or possibly absent, mesocyclone classification is displayed as a result. This creates a feature known as the "cone of silence" on the radar screen.

Mesocyclones should not be considered a formation that is viewable, as tornadoes are. Some cloud features may suggest that a mesocyclone is present in a thunderstorm, such as curved feeder or inflow bands.

Do not question the judgment of an expert National Weather Service employee if a warning has been issued for your area.

The probability of severe hail, which is hail over three-quarter inch in diameter, is known as POSH, and is displayed as a percentage within the POSH column on the storm attribute table.

The probability of hail, which could be any size, is known is POH, and is also displayed as a percentage on the storm attribute table.

Both POSH and POH are determined for each thunderstorm on radar as a whole, given they have a cell identification code.

If the radar is unable to determine the probability of hail or severe hail, the abbreviation "Unkn" will be displayed. "Unkn" stands for unknown. Cells a great distance from the radar site, as well as those which have just developed, may have an unknown, or undetectable, probability. Check the next scan as to if the unknown probability has been resolved, as it may. If not, try using a different radar site that the storm cell may be closer to. Thunderstorm cells that have an unknown probability also have an unknown maximum hail size. Either all hailstone-related information is present, or none of it is.

National Weather Service radars also plot a green triangle on the radar screen for a storm cell containing hail under special circumstances. The triangle is unfilled or filled depending on the intensity, but not plotted where the largest hail is expected to fall. An unfilled triangle indicates the probability of severe hail (POSH) is 30 or 40 percent. When the radar detects a 50 percent chance of severe hail or greater, the triangle becomes filled. Not all radar viewers show this hail triangle, including StormLab©, as it would clutter the radar screen during a severe thunderstorm outbreak. Nonetheless, StormLab© does provide the user with the option to set audio and visual alarms for thresholds of several radar-estimated storm cell characteristics.

Hail algorithms are basic given the radar is supplied with the height of the freezing level (0° C) and the placement of the -20° C layer in the atmosphere. Therefore, the hail probabilities and estimates can be misinterpreted by a user who does not have a great deal of experience.

Never doubt or question an expert as to why a warning was or was not issued for your area, even if there was a 100% severe probability of one-inch diameter hail. Different atmospheric conditions can alter the size of hail. As a result, radar operators sometimes have their own techniques to determine hailstone size and probabilities.

Needless to say, if radar detects a 100% chance of severe hail to 3.25 inches in diameter, it would probably be a good idea to seek safe shelter.

Look for the largest hailstones and best chances of it under areas of high reflectivity on the composite radar scan.

A radar cannot detect the probability of general hail and severe hail well, or possibly at all, if a thunderstorm is close to the radar site. Beams are unable to extend into the updraft at such a degree from the radar's transmitter. A lower, or possibly absent, hail probability is displayed as a result. This creates a feature known as the "cone of silence" on the radar screen.

Radar also tries to determine the maximum diameter of hail falling within a storm cell. The output from this is formatted into inches and placed in the "Max Size" column of the storm attribute table.

The size that is estimated by radar is usually overdone and the majority of hail that falls is generally smaller. This is especially true during the summer months. Therefore, National Weather Service radar operators tend to only use the maximum hail size to issue warnings when the probability of severe hail (POSH) is over 50%.

Radar hailstone estimates range from <.50 inch to >4.00 inches in diameter by quarter (.25) increments. StormLab© shades the table cell of the maximum hailstone size pink once the estimated size exceeds four inches. Other shadings occur based on alarms set by the user through the Storm Cell Alert feature.

If the radar is unable to determine the maximum size of a hailstone, the abbreviation "Unkn" will be displayed. "Unkn" stands for unknown. Cells a great distance from the radar site, as well as those which have just developed, may have an unknown, or undetectable, hail size. Check the next scan as to if the unknown size has been resolved, as it may have. If not, try using a different radar site that the storm cell may be closer to. If a maximum hailstone size is not detectable, or unknown, the probabilities of hail and severe hail will not be undetectable as well.

A radar cannot detect the maximum size of hail well, or possibly at all, if a thunderstorm is close to the radar site. Beams are unable to extend into the updraft at such a degree from the radar's transmitter. A lower, or possibly absent, maximum hail size is displayed as a result. This creates a feature known as the "cone of silence" on the radar screen.

Never question the decision of an expert meteorologist to issue or withhold a warning.

Unlike the VIL product, which determines vertically integrated liquid using grid boxes (also known as grid-based VIL), the VIL number used on the storm attribute table is cell-based. This allows for the vertically integrated liquid associated with composite reflectivity to read slightly higher than that of the VIL product, especially when a thunderstorm is tilted highly.

The reason for the difference is the way the reading is taken. With cell-based VIL, the radar's elevation scans slice the storm nearly horizontally, searching for the highest possible reading. With grid-based VIL, however, the radar's measurements are made in vertical grid VIL boxes. Therefore, with grid-based VIL, the highest levels of vertically integrated liquid are more apt to being overlooked.

Either way, vertically integrated liquid is the amount of liquid in a vertical column. The output is displayed in kilograms per square meter.

Radar operators often use VIL to identify storms with heavy rainfall and large hail. It can also be used to predict the onset of wind damage if it is combined with other products.

There is not a set VIL number which determines when storms are capable of producing large hail. The reason for this is due to the freezing level in the atmosphere and the height of the updraft. In general, the layer of freezing can be found higher in the atmosphere during summer, and lower, possibly at the surface, during winter. The lower the freezing layer is in the updraft, the better chance there is that hail can be expected to fall on the surface. As the freezing layer becomes higher in the atmosphere, a larger updraft tower is required for hail, and therefore a heavy rain event is more likely to occur.

A frontal boundary, such as a cold front, can also have an affect on how VIL numbers correlate to the size of hail. Behind a front, when cold air is aloft, a VIL of 10 to 25 is sufficient to sustain nickel-sized hail. However, warm air aloft in front of a cold front can cause storms with a VIL of 25 to 40 to only drop dime-sized hail.

During the summer, depending on your location, a high vertically integrated liquid number, one that favors heavy rainfall or large hail, is usually between 50 and 65. This number is typically less during the winter months. A Storm Cell Alert feature is included with StormLab© that sounds an alarm if the VIL of a certain cell meets or exceeds the red criteria set by the user, given the alert is on. The preset value for an alarm is a VIL of 55.

A radar cannot detect vertically integrated liquid well, or possibly at all, if a thunderstorm is close to the radar site. Beams are unable to extend into the updraft at such a degree from the radar's transmitter. Limited, if any, VIL is displayed as a result. This creates a feature known as the "cone of silence" on the radar screen.

The dBZ Max on the storm attribute table is the highest reflectivity found within a particular thunderstorm cell. Depending on the reflectivity, a radar operator can determine the rate at which rain is falling as well as the potential for severe hail.

Rain starts to fall when the reflectivity (dBZ) is around 20. Of course, a set of atmospheric conditions can alter this. A very moist and saturated environment could allow rain or drizzle at a lesser reflectivity level. However, a dry environment could allow rain to reach the surface only at a higher reflectivity, as some of the precipitation could evaporate before it hits the ground.

Snow falls at much lower reflectivity levels. The radar will likely not tag even a band of heavy snow, unless it is associated with convection, creating thundersnow, for the storm attribute table. In general, snow reflectivities are less than 30 dBZ.

A reflectivity (dBZ) of around 55 may suggest large hail is falling, but not always. For a better indication as to if hail is falling with a storm, compare the freezing level in the atmosphere to the size of the updraft tower. The lower the freezing level intersects with the updraft, the better chance there is of hail, given the updraft tower has been sustained. Referring to the vertically integrated liquid, or VIL, may also be a good idea.

On the radar image, one indication of rather large hail with a cell is a hail spike. A hail spike is shown extending radially outward from a radar site (pointing the opposite direction) when the radar beam crosses through the core of a thunderstorm.

The following table shows the potential rainfall rate per hour at select reflectivity (dBZ) rates.

This set of rainfall estimates per dBZ level is variable, and can alter from time to time or from site to site, depending on other conditions, such as environmental saturation. For example, hail and sleet often reflect the radar beams better than rain. This allows a higher dBZ to be present in those areas on radar. It is not true rainfall, so the chart has to be modified. Radar operators often take steps to assure that high dBZ levels, which are actually hail or sleet, are not misinterpreted as heavy rainfall.

StormLab© allows users to create their own color schemes for viewing reflectivity imagery using the Color Table Editor. Unique data sampling in StormLab© also allows a user to place his/her computer mouse over reflectivity to determine where it fits on the key on the right side of the image.

A radar cannot detect the maximum reflectivity (dBZ) well, or possibly at all, if a thunderstorm is close to the radar site. Beams are unable to extend into the updraft at such a degree from the radar's transmitter. A lower maximum reflectivity is displayed as a result, given if the reflectivity is absent, the cell would no longer appear on the radar image. This creates a feature known as the "cone of silence" on the radar screen.

The vertical height of the maximum reflectivity (dBZ) can also be determined.

The "@ Height" column in the storm attribute table corresponds to the vertical height of the maximum reflectivity (dBZ) for a certain thunderstorm cell. The output should normally be a four-digit or five-digit number displayed in feet above radar level (ARL).

The National Weather Service radars originally display a height in composite reflectivity as a one-digit or two-digit number with a one-digit decimal. Multiply this number by one thousand to correctly format the output. For example, if a radar displays the height as 15.6, the actual height is 15,600. StormLab's© storm attribute table does the math for you by showing the actual height directly upon loading.

Radar operators use the reflectivity height to determine the structure of a thunderstorm, and where the most intense portion of the updraft is. The reflectivity height may also be combined with other products to form an algorithm regarding the potential for damaging downburst winds. Those in the aviation industry may also find the height of the maximum dBZ useful so pilots can avoid flying planes into extreme turbulence. In general though, aviators attempt to avoid strong and severe thunderstorms, as well as give them a decent amount of room.

A radar cannot detect the maximum reflectivity (dBZ) height well, or possibly at all, if a thunderstorm is close to the radar site. Beams are unable to extend into the updraft at such a degree from the radar's transmitter. A lower maximum reflectivity height is displayed as a result, given if the reflectivity is absent, it would not have a height, and the cell would no longer appear on the radar image. This creates a feature known as the "cone of silence" on the radar screen.

The "Top" column in the storm attribute table corresponds to the highest vertical height of a 30 dBZ echo for a certain thunderstorm cell. The output should normally be a four-digit or five-digit number displayed in feet above radar level (ARL).

Depending on the distance from the radar site, and the tilt of the beams, the echo top can be greatly overestimated or underestimated to the order of five thousand to ten thousand feet.

The radar may not always be able to resolve the 30 dBZ echo top. When this error occurs, the output reading will have a "<" or ">" in front of it, meaning the top is less than or greater than, respectively, the number that follows. For example, if the reading <26,800 is shown, then the top is less than 26,800 feet.

The storm attribute table's "Top" column does not produce the same readings that are on the "Echo Tops" product because the product is not reliant on the 30 dBZ echo, but the column's output is. In addition, the product is generally less precise than the column. The reason for this lies on the fact that the algorithm associated with the "Echo Tops" product measures the echo top with the radar beam that is closest to it. At far distances, it can appear on the product that the height of the clouds sharply varies between two levels five thousand feet apart, whereas the variance between tops is more gradual.

A radar cannot detect the highest vertical 30 dBZ height well if a thunderstorm is close to the radar site. Beams are unable to extend into the updraft at such a degree from the radar's transmitter. A lower 30 dBZ height is displayed as a result, given if the vertical height is absent, the cell would no longer appear on the radar image. This creates a feature known as the "cone of silence" on the radar screen.

This column on the storm attribute table lists the forecasted movement of each thunderstorm cell. On the radar image itself, a projected path is displayed starting at the cell identification code, and extending outward.

The forecasted movement of any given thunderstorm cell is separated into two parts within the column. On the left, the first part is the direction the movement is from, in degrees, which is inverse from where the storm is headed. On the right, the second part consists of the speed, in knots, at which the storm is moving. For an example, if the output of one thunderstorm in the "Fcst Mvmt" column was 270/33, the cell would be moving toward the east at 33 knots.

The movement is calculated from the centeroid of each storm cell, and uses each cell's history. A centeroid can be determined from the "Az/Ran" output, which is also plotted on the radar image, if desired by the user, and is the base of the projected path.

The length of the projected path depends on the speed at which the storm cell is moving. It is consistent of five points. The first point indicates where the cell was last detected, with the following four points representing fifteen-minute tracking intervals up to one hour. For example, a particular cell is expected be at the third point in thirty minutes, and the fifth one in about an hour, according to the radar.

In the event that a storm cell has just developed, and is identified by radar, the word "New" may be displayed for the forecasted movement because not enough data is present for the radar's algorithm to detect where it is headed. Developing thunderstorms can often have unusual paths since the radar's algorithm sometimes misinterprets cell development as movement.

Squall lines may have several forecasted paths. With luck, they will all be moving at around the same pace in the same general direction. If not, take an average. When dealing with the evolution of a bow echo, remember that the tail will often extend well outward from the head as it grows. As a result, the projected path of the head may be slightly slower than the tail, as well as deviate from the general direction of the rest of the bow or comma echo by up to 45 degrees.

As briefly mentioned above, it is a very good practice to take an average of the various speeds and directions coming from the different storm cells on radar. Thunderstorm cells that have been in existence the longest tend to have the best projected paths. Try to rule out any odd and abnormal paths if possible. Odd paths are often adjusted by the radar between runs to be more accurate. Be careful though, as cells moving in front of a squall line can often deviate from the movement of the line. This situation, along with several others, makes the job of radar operators difficult at times, especially when they are trying to issue advanced storm warnings. Nonetheless, always heed National Weather Service warnings, which are issued by experts, to seek shelter immediately.

Cells that are within a few miles of the radar site may appear weaker than they actually are. This is because the beams of the radar are unable to tilt at such a high angle to detect the full reflectivity of the storm. Due to the "cone of silence," an abnormal or odd projection path could result if the cell has not been robbed of the tracking code it once held. Always leave the potential open that a thunderstorm could be weakening however.

StormLab© uses data from the "Fcst Mvmt" column, along with an extensive database of cities and towns throughout the United States, to generate the StormTracker Plus™ Pathcast included with the professional version. This helps users to determine when a dangerous storm will reach their location right down to the minute!

Anomalous Propagation, or AP, is caused by the ground when a radar's beam is bent and slants into the surface at rather large distances from the radar site. AP is usually seen as an area of mixed dBZ feedback on the lower reflectivity scans, and has no logic or organization to it as a thunderstorm cell does. Base velocities are usually around zero for true AP echoes.

Since National Weather Service radars do not sample the entire volume of air, but instead just special slices, radar output from a single storm can change from scan to scan as it moves from one slice or tilt into another.

At around twenty-four nautical miles from the radar site, everything above fifty thousand feet goes undetected. This is hardly noticeable unless the storms in the area are well sustained.

When a storm tracks within about fifteen nautical miles of a radar site, the "cone of silence" effect occurs because everything above thirty thousand feet goes undetected. The "cone of silence" is caused when radar beams are unable to tilt high enough to intercept a storm nearby, or overhead, the radar site. As a result, the storm passes undetected by the radar, and is therefore "silent."

In short, the "cone of silence" is an area of the atmosphere which cannot be sampled by the radar.

The "cone of silence" effect is very noticeable at six nautical miles or less from the radar site, where everything above ten thousand feet goes undetected.

Even though storms can weaken, when they near a radar site, the "cone of silence" effect is likely occurring. Just to be sure that is the case, check the atmospheric conditions, or retrieve radar images from a nearby location.

Note: One nautical mile is equal to 6076.115 feet, or 1852 meters, in the United States.

A special thanks goes to those who have provided valuable information on reading composite radar imagery. Please visit their web sites for more excellent content relating to weather radar.

dBZ	Rate
15	.01"
20	.02"
25	.04"
30	.09"
35	.21"
40	.48"
45	1.10"
50	2.50"
55	5.68"
60	12.93"
65	16+"