The first Intensive Observing Period (IOP) of VORTEX-SE is completed.
Observations began in the early afternoon on Sunday, 13 March, with a sounding launch by NOAA’s ARL/ATDD from the SWIRLL facility at the University of Alabama in Huntsville (UAH). UAH is hosting VORTEX-SE, and the Operations Center is at the SWIRLL.
Shortly after the midday briefing and weather discussion, the lead investigators in VORTEX-SE huddled to come up with the science plan for this episode. The main weather feature of concern was a compact, fast-moving upper low pressure system moving out of southeast Oklahoma, expected to cross Arkansas and the Memphis, TN area during the nighttime and early morning hours. We were fairly sure a system this intense would be associated with thunderstorms. Much of the forecast guidance indicated that, in northern Alabama, there would be little or no surface-based CAPE (a measure of the energy available to produce buoyant, rising motion in air that is lifted from the ground feeding a thunderstorm updraft). This doesn’t mean no storms… thunderstorms can be energized by air that originates somewhere above the ground… but tornadoes are unlikely when the surface-based CAPE is zero. So our decision was to deploy as far west in northern Alabama as we could, hoping that the storms would still be somewhat energetic and perhaps there would be a little lingering surface-based CAPE before nighttime cooling removed it completely. Where we can deploy is largely determined by where we can find suitable sites for mobile Doppler radars: because so many crucial processes in tornado formation happen near the ground, this requires a good radar view of the horizon without hills, trees, power lines, buildings, etc. obstructing the view.
Most of the research teams had found locations for their measurements by evening. Sounding teams were arrayed along a line from Huntsville toward the west-northwest into northeastern Mississippi. Mobile Doppler radars were arrayed in a triangle near Florence, AL, and other instruments joined them there, such as deployable disdrometers to measure raindrop sizes and numbers, and deployable “sticknets” to measure the temperature, humidity, pressure, and wind near the ground.
During the night, it became apparent from our soundings and the surface data that in fact the near-ground air had cooled under nearly clear skies, eliminating all surface-based CAPE. Nevertheless, the band of thunderstorm activity continued to advance east across MS, and into AL around 2 AM. With these storms, there was only very weak horizontal shear (change in wind from one area to another, about a mile apart, that might be a precursor to vortex formation) on a couple of occasions. The storms were feeding off of air from well above the ground, and in this case no severe weather was reported at the ground. By around 5 AM, the storms had weakened to showers and showed signs of departing the area to the north, and our observing operations were ended.
Was the mission a success? I have some thoughts coming together that I will try to write down in a blog post in the “Perspectives” category soon. We failed to observe any low-level vortices that could have developed into tornadoes, but I think we were successful in several ways.
First, one of the key aspects of VORTEX-SE is learning how to make observations to answer important tornado questions in the Southeast. This is far from trivial. We have no mobility in the traditional sense of storm chasing: hills, trees, low clouds, sparse winding roads, rapid storm motion, storms at night, lack of deployment sites, and other factors make mobility nearly impossible. So the challenge becomes how to forecast the probable weather with enough lead time to put together a scientifically useful observing network. This means anticipating the structure, motion, and intensity of storms at least a few hours before they occur, instead of on the time scales of minutes that we became accustomed to in earlier tornado research programs. In developing the network, important considerations are Doppler radar placement to allow for overlapping beams (giving us the ability to estimate 3D wind), and focusing the other observing systems within those multiple-radar networks. In this first IOP, we were able to iron out the many of the wrinkles in deployment strategies, communication, and coordination.
Second, these “null cases” where we fail to observe tornadoes are important. This is under-appreciated, even among many tornado researchers. Thinking back across many years of storm chasing in the Plains, one of the toughest forecasts to make is when there is huge CAPE, but the warm, humid low-level air is trapped below a warm, stable layer of air aloft… the so-called “cap”. We can compute a temperature called the “convective temperature” at which the air near the ground becomes warm enough to remove that stable layer, and convection ensues (it’s not that simple, but in broad terms the concept works pretty well). So all this available energy can be associated with no deep cumulus clouds, or it can be associated with monster supercells, depending on a few degrees Fahrenheit difference in temperature near the ground. Forecasting whether that “tipping point” occurs can be a big challenge with big consequences. A similar situation exists in the Southeast: a few degrees difference in low-level temperature (and similarly, small changes in humidity) can mean the difference between no surface-based CAPE with small chances of vortex development near the ground, or “some” surface-based CAPE and much greater chances of vortex development. To be clear, surface-based CAPE is just one number that we use to get a sense of low-level conditions; the intensification of a vortex near the ground is too complex to be understood using these simple parameters alone. But no matter how one characterizes the near-ground conditions, it is likely that subtle differences can change the atmosphere from suppressing low-level vortex formation to favoring it. In an ideal (lucky) IOP, we would observe a case where part of the domain had just-good-enough air for low-level spinups, and part of the area did not, and then we could compare the resulting weather in those two areas. In the IOP we just finished, we were successful in sampling conditions that were slightly on the unfavorable side.
In VORTEX-SE we hope to make observations like these 4-5 times during March and April. I hope we can observe a variety of regimes that have tornado potential during the remainder of our IOPs.