Detecting Tornadoes


Violent tornadoes like the one that leveled parts of Moore, Okla., a few weeks ago momentarily bring severe weather into the public consciousness. The now common images of the churning, chaotic, and massive vortex bearing down on a city are among the most frightening imaginable. For the people in the path, there is nothing to be done but hide.

And the images taken afterwards are just as incredible. Rows of sturdy houses wiped off of their foundations, leaving nothing but concrete slabs – and the obvious fact that anyone inside and above ground had little chance of survival.

While these tornadoes may seem like events from a world out of control, in which people are powerless in the face of a hostile environment, this is not at all the case. This tornado was clearly visible to meteorologists before it even touched the ground or caused a single dollar of damage; and as a result, tornado warnings were issued and broadcast, probably saving thousands of lives.

More than 35 minutes before Moore was hit, the National Weather Service Doppler radar station in Norman, Okla., first detected rotating winds near the base of the cloud. A tornado warning was issued, and the sirens began sounding in Newcastle, Moore, and surrounding cities. Figure 1 shows a radar image taken at the point of the tornado’s maximum strength, just before it entered the Moore area. These images – and specifically the right side, showing the storm’s relative wind velocity – are the primary justification for most tornado warnings. The radar can see the signature of a developing tornado well before it descends from the cloud.

Tornadoes are usually produced by supercells, a class of thunderstorms characterized by powerful rotating updrafts. The key requirement for these storms is wind shear, or winds that change speed and direction with height. For instance, the winds aloft on a day with strong shear might be something like this: 40 mph from the south at 3,000 feet, 70 mph from the southwest at 10,000 feet, and 120 mph from the west at 15,000 feet.

All thunderstorms have strong updrafts that carry warm, moist air from the surface high into the atmosphere, where clouds and rain condense out. In a highly sheared environment, that updraft can begin to rotate – and a tornado occurs when the rotating updraft extends down to the surface. However, what exactly causes a tornado to form in a particular storm is still not well understood.

In addition, the data available in real time is usually limited to radar images. While radar can reliably detect thunderstorm rotation, only about 25 percent of radar-detected rotation events ultimately result in tornado touchdowns. There is currently no way to distinguish between rotation that will lead to a tornado and that which will not, and so there is little choice but to issue a tornado warning for each rotating storm.

While this method results in the vast majority of tornadoes receiving official warnings, it also means that the tornado warning false alarm rate hovers between 75 and 80 percent. This is a major problem for the tornado forecasting community, and one that has proved difficult to address. Improvements in the false alarm rate will likely come as the capabilities of radar systems increase and as more advanced radar data processing techniques are adopted.

Despite the challenges, tornado forecasting today is extraordinarily good. Chances are, if you are hit by a tornado, you will have been warned. The two key factors in reducing the fatality rate are ensuring that all people have access to underground tornado shelters, and convincing people to seek shelter during every warning – as is often the case, the hardest problems are social rather than technical.

In many ways, these severe storms have been tamed. The ability to give the public accurate tornado warnings with sufficient lead-time is a significant achievement, and was simply impossible not too long ago. As tragic as the tornado in Moore was, the fact that only twenty-four people were killed when an EF-5 moved through an urban area is a victory for the scientific community, albeit one that can certainly be improved upon.

In the image: Left: base reflectivity, showing precipitation intensity (red/pink is most intense). The “hook” shape of the storm is caused by rain and hail wrapping around the rotating updraft at the center of the hook. The large pink area just southwest of Moore shows the dense debris cloud around the tornado. Right: storm relative wind velocity. Green represents wind moving towards the radar, and red is wind moving away. The area where bright red meets bright green is an indication of rotation, and in this case, a powerful tornado already on the ground. The radar image was created using the GRLevel3 radar analysis tool.



Add new comment

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd><p>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Refresh Type the characters you see in this picture. Type the characters you see in the picture; if you can't read them, submit the form and a new image will be generated. Not case sensitive.  Switch to audio verification.