Meteorologists are often tasked with making predictions, but understanding the present can be equally important and similarly daunting. When faced with tornadoes, which can develop, intensify and cause extreme destruction in a matter of minutes, such “nowcasting” becomes crucial.
Radar is a powerful tool for meteorologists aiming to understand the active state of the atmosphere. Beams of microwave energy emitted by one of 160 WSR88-D NEXRAD Doppler stations in the United States bounce off airborne objects, returning to sensors that can infer their speed and size. These Doppler radars are the primary sensors among additional smaller-scale radars across the country, giving the United States by far the most weather radar capability in the world.
Because most of the bodies in the atmosphere are water in some form — rain, snow, sleet, hail, etc. — radar is useful for detecting precipitation.
But the ability of radar to figure out wind speed and the size of objects in the atmosphere also makes it possible for meteorologists to detect tornadoes, and even estimate their intensity, in real time. This is done daily by the National Weather Service and is how the agency generally knows to issue a tornado warning.
Civilians have access to much of the same radar data used by the Weather Service, which makes it possible for weather enthusiasts to follow along from home.
When people think of radar, they probably picture reflectivity. This basic output shows meteorologists the density of airborne objects in a certain volume, which makes it useful for estimating the type and intensity of precipitation. This output is most frequently shown on television and in weather apps.
A hook echo is a type of storm structure on reflectivity radar that shows the storm is rotating, and may produce a tornado. Ideally, such an echo looks like a spiral turning outward in a clockwise way, with the “thickness” of the precipitation increasing — or, if you will, a hook shape. The tornado is found at the spiral’s narrow apex.
Not all hook echoes will look so clean. Sometimes, the hook is embedded in a line of storms or is barely visible under a huge supercell. Sometimes — especially in the Mid-South — there is persistent indentation or notch in the larger cell. Other times, a storm is too far away from radar for the hook to be clearly evident.
A hook echo tells meteorologists that a storm is structured in such a way that a tornado may be imminent. But a far more important piece of the puzzle is wind speed and direction, which radar also can provide.
A radar product that many are probably less familiar with is velocity data, which indicates the speed and direction of airborne objects — or rain — detected by the technology. This type of radar data can show observers where rotation may be occurring within storms. Such rotation is a prerequisite for tornado formation.
Generally, velocity data is shown in green where winds are blowing toward the radar, and red where they are blowing away. When rotation is occurring, winds blow toward the radar and away from the radar in proximity — in most cases, red and green must touch. This is called a couplet.
Rotation is easy to confuse with wind that is moving together or spreading apart, which also involve couplets of touching red and green. Therefore, it’s important to note the orientation of the couplet relative to the radar. If a line is drawn from the touching red and green to the radar, a rotational couplet will have maximum velocity values moving outward from the line, while a divergent or convergent couplet will have maximum velocity values moving along the line.
To make sure that the rotation seen is real and not simply a product of the radar, a final check is to ensure that the velocity couplet overlaps with reflectivity returns. Ideally, the touching green and red velocity data should overlap with the hook echo’s apex or within certain areas of linear storms that are less noticeable on reflectivity.
If a velocity couplet is found and a rotational origin that overlaps with a hook echo is confirmed, an observer can be confident that the storm is rotating. The more “tightly wound” the rotation, and the stronger (or brighter) the involved velocity, the greater the likelihood that a tornado is on the ground.
It’s important to note that in many instances, the velocity couplet is an indicator of parent rotation within the supercell. This larger vortex is termed the mesocyclone, and it coincides with the storm’s updraft. Not all mesocyclones generate tornadoes, and when a tornado forms, it is generally a much smaller, intense vortex contained within the larger mesocyclone. Depending on how close the radar is to the supercell, the “tornado vortex signature” as it is called, may not be resolved.
The stronger a tornado, the faster the winds blowing around it.
Velocity data is a crucial indicator that a tornado could be imminent, but it cannot inform observers of the actual presence of a tornado on the ground. Luckily, there are other radar tools that can.
When a tornado is on the ground and doing damage, it lofts dirt, plant matter and other small debris high into the atmosphere. Because radar is designed to detect the presence of airborne objects, it can show meteorologists where debris is present in the atmosphere.
When no storm spotters are present to make direct observations, recognizing debris on radar can be crucial for saying with confidence that a tornado is on the ground. Without a direct observation, even a strong velocity couplet and a well-organized hook echo can indicate only the likelihood of tornado development, while lofted debris can reasonably confirm the presence of a damaging twister. (There are cases where debris is present without a tornado, but those instances are not common.)
On reflectivity imagery, debris generally looks like a “ball” of high values near the apex of the hook. These high-value balls are fairly telling, but confirming the presence of debris requires an additional step.
When objects are lofted into the atmosphere, modern radar can go beyond just measuring density. A tool known as correlation coefficient shows the amount that airborne bodies in the radar beam’s path differ from one another. Rain drops tend to be very similar in size and shape, while debris is not, meaning correlation coefficient can effectively differentiate between the two.
On radar imagery, where correlation coefficient is noticeably low, something that is not precipitation is likely to be present in the atmosphere. Where low correlation coefficient overlaps with a reflectivity ball and a strong rotation couplet in velocity data, a tornado is almost certainly doing damage. When this happens, the Weather Service will add “radar-confirmed tornado” in the text of the tornado warning.
The stronger a tornado, the higher it can lift debris. Meteorologists can therefore use correlation coefficient to help determine not only the presence of a tornado, but also a twister’s potential intensity. Nowcasting intensity is an evolving science, with storm surveys the primary vehicle to determine rating after the fact.
Certain apps provide higher-resolution data, although many cost money or require some technical know-how to use. Radarscope, a phone app, is generally the most popular among meteorologists and committed enthusiasts, although the app costs around $10. Other options include Radar Omega and MyRadar. GR2 analyst, a computer application, also provides very high-quality data but is quite expensive for the general hobbyist.