Tornados kill hundreds of Americans a year, leaving devastation in their wake, and yet they are still poorly understood. We speak to the people who chase them in the name of science.
Tornados and thunderstorms are the most powerful winds on the planet, leaving devastation across north west America on average six times per year. Yet, despite the finest minds and the most sophisticated technology being applied to the science of severe weather, the optimum conditions for the evolution of a tornado remain largely a mystery.
"Our ultimate goal is to improve warning times on tornados," says Joshua Wurman, CEO at the Centre for Severe Weather Research (CFSWR). "But our other equally important goal is to understand the structure of evolving tornados. Surprisingly little is known about the structure of a tornado and how strong the winds really are. When we finally understand how a tornado forms, we can begin to engineer houses and cities to better survive them."
Tornado warning times have improved drastically over the last 25 years. Previously, inhabitants of tornado-struck cities had approximately three minutes' lead time, with 90 per cent of all warnings proving to be false alarms. Now the advance warning period has increased to 13-15 minutes, meaning more evacuees are able to reach storm shelters. The false alarm rate is still fairly high, at 70 per cent, which Wurman says can cause a dangerous delay in reaction time, as evacuees begin to believe every situation will be a false alarm.
He says determining specifically how a tornado forms is on a par with the medical profession's understanding of cancer formation; there's still a long way to go before researchers can to get to the stage where a useful prediction can be made into why an individual situation does or doesn't progress.
"We know the broad strokes of how tornados form, but we still don't know the finer details, which is why it's still such an active area of study."
What meteorologists are aware of are the basic conditions for the generation of the parent-storm which they call a 'supercell'. They can predict fairly accurately which days will yield a supercell storm, and have determined that the average tornado conversion rate is around 25 per cent. Currently, what they are unable to determine is why the remaining 75 per cent don't become tornados.
The basic setting for a tornado is that, temperature and humidity permitting, winds that are close to the ground and moving slower than those higher up cause the air-mass of a supercell to lift and rotate horizontally like a car axle. The interaction of resulting up- and down-drafts tilts the air mass up on to one end. As long as the air pressure is low, the spinning cloud will reach the ground via the path of least resistance, creating a tornado.
"We can't tell which storms will turn into tornados, or which ones will make strong tornados or weak ones. We can't even tell when, during the lifecycle of a storm, it will become a tornado." says Wurman. Needless to say, this is not exactly comforting news for the 32 million people who reside in the states of Texas, Kansas and Oklahoma, America's infamous 'Tornado Alley'.
However, sophisticated new test and measurement tools are enabling researchers to gain fresh insights into tornado behaviour. Although conceptual models still make up a significant portion of severe-weather intelligence, engineers can produce more accurate simulations through the use of meteorologist-written software. Wurman says that current computational modelling can only be as accurate as the quality of the data gathered in the field.
"Our state-of-the-art computational modelling processes have improved in recent years, but there is still a long way to go to make realistic tornado simulations," he says. "There are some models that never generate a tornado and there are some that always generate a tornado, which is nice, but not accurate and not entirely true. Ideally in 20-30 years we'd like to have computational models that can look at a thunderstorm at zero minutes and be able to predict which will make tornados and which won't. The only real way to find out what is going on inside these storms is through direct observation."
Eye of the storm
There are essentially two measurements to take in any given tornado; those surrounding the thunderstorm, and those at its heart. To measure the external conditions, research teams drive within a 1-2km radius of the storm in reinforced pick-up trucks with mesonet and mobile radar systems called Dopplers (DOW) mounted to the front,costing around $1m each. DOW systems measure wind speed, direction, humidity and static pressure from the surrounding storm which is data-logged onto the team's server. Analogue radars scan pulses through the storm in 4D, gathering spatial and temporal information, a tool critical to informed external measurements.
If created quickly enough, the 3D images of wind pattern, precipitation fields and hail patterns are produced to show 4D animations of the spatial and temporal evolution of each field. The radar acts like an HD microscope, picking up miniscule detail tens of thousands of times more accurately than the human eye. To measure wind speed, blade anemomenters are pole-mounted to the trucks, along with ultrasonic anemometers, which send out pulses of sound. Up to eight channels of radar are transmitted back to analogue receivers, which convert the voltage frequencies down into lower frequencies of around 100MHz, which is then processed and saved as raw data.
Measurements such as these generate a huge amount of data, so Tim Samaras, severe weather researcher for TWISTEX (Tactical Weather Instrumented Sampling in/near Tornadoes EXperiment) uses a National Instruments industrial controller and CompactDAQ and DIAdem for data acquisition and storage, which in the field of storm measurement can often reach a terabyte a day. He has designed a bespoke virtual instrument - a quick-look dashboard – which allows him to sync and visually represent all streams of this data. A LabVIEW interface analyses the static-pressure data that has been collected; the lower the static pressure, the more powerful the tornado will be as the wind attempts to rebalance the dropping pressure in the air. Scientists use these biometric pressure readings to try and determine how the wind speed influences the frequency and speed of a tornado from inside the wind tunnel.
But the most dangerous, and therefore most complex, measurements to take are those taken from within the heart of the tornado. The danger associated with this form of research is so severe that it has resulted in Samaras becoming a star in his own right for the Discovery Channel's documentary series, 'Storm Chasers'. "Our primary research is to collect data from core tornado environments to better understand tornado genesis, decay and dynamics," says Samaras. "This means looking at wind, temperature and humidity shift boundaries associated with supercells right at the heart of the tornado."
Probing for information
Storm chasers have come up with a relatively crude engineering solution for internal measurement, or at least crude in comparison to the high-tech solutions they utilise across other areas of severe weather research.
With winds reaching 300mph and above, no rugged vehicle can withstand the internal conditions of a tornado, let alone a human complete with sensitive measurement equipment. Instead, 50kg rugged HITPR probes or 'tornado pods' are deployed. Different teams use bespoke pods and technologies, but normally sensory equipment is attached to a reinforced steel shell which protects a silicon-insulated 'black box' data-logger, and placed directly in the path of an approaching tornado. This sensory equipment includes instruments such as modified pitot-static anemometers with embedded processors – normally employed on airport runways to measure aircraft speed.
The probes, which are capable of withstanding wind speeds of 300mph, measure static pressure from within the tornado at three different heights ranging from 0.75m to 2m high. Although the auxiliary sensory equipment is normally destroyed, the probes continue to collect and transmit four channels of audio, temperature, wind speed and directional data for as long as possible from inside the tornado back to an external hard-drive. Some probes also feature video recording equipment, which provides visual feedback showing what dictates the interior structure of a resulting tornado.
The missing piece of the puzzle for storm chasers is an abundance of detailed data on relative temperature and humidity from inside the storm, which to provide relevant analysis needs temporal sensitivity of within one or two degrees centigrade. Temperature and humidity can be measured on a remote basis via a mobile mesonet system, but only from the exterior of the storm and at low level says Wurman. "The radars can't really give us relative temperature or humidity readings. All they're measuring is wind, and the location of the rain and hail."
Although HITPR probes provide good levels of measurement from ground level, the next step for storm chasers is to utilise a technology which will allow them to measure inside the storm but 2,000m into the air. University of Colorado, in partnership with the CFSWR, has pioneered the use of unmanned air vehicles (UAVs) alongside 80 other weather-measurement instruments in a $12m project called Vortex 2. Despite their innovation, the UAVs were unable collect quality data from the inside the storm due to the violent conditions. Currently the only successful internal measurement devices are weather balloons launched by hand, but they cannot be deployed into a specific area of the storm and are only able to record one line of data at a time.
By choice, storm chasers prefer to operate in the western area of tornado alley due to the landscape. Fewer trees and gradient on flatter terrain mean there is less interference with test and measurement instruments, and it is easier to escape from a tornado should a storm begin to turn. Wurman and his team have only been hit by a violent tornado twice; good odds considering they have collected data from over 170 tornados. Their closest call was in Geary, Oklahoma in May 2004.
The tornado came within a kilometre of where the team was stationed, yet when they began to speed away from the storm, they were intercepted by winds stronger than those prevalent in the tornado itself. "We couldn't drive in it. I was yelling to the driver to go faster and he told me his accelerator was pushed all the way to the floor, yet we were still stopped dead on the highway."
External winds of over 300kmph prevented the team from moving away from the tornado and obscured the vision of the drivers, whipping the surrounding debris up to create an opaque field of vision. "We were worried that despite what our radar technology was telling us, we were driving into the tornado rather than away from it. It was terrifying, because we could not understand why we were trapped in such strong winds."
Luckily, at the last minute the tornado veered away from the trucks, taking only the door of one of the cabins and several antennas with it. "The important thing is that we continued to gather data – our antennas kept running until the last minute. This is probably the moment when our lives should have been flashing in front of our eyes, but we were more concerned with keeping our computers running."
Samaras, who has suffered similarly terrifying experiences all in the name of research believes the way to avoid civilians experiencing these types of encounters in the future is through numerical modelling. "The future of severe weather research lies in numerical modelling, the improved ability to accurately forecast the weather and predict accurately where a tornado will come to and the direction it has come from." He believes it could be 30 years before the secrets of the tornado are fully realised.
"Some scientists have a theory that tornados can approach sonic level," he says, "I want to record this level of sound which will potentially generate astronomical amounts of data. It will generate so much data that it will just keep recording until the HD card is full." *