Wednesday, July 28, 2010

The Great Tornado Chase

Texas Tech researchers work to solve the mystery of the tornado after a 6-week journey across the Great Plains.

2010 Texas Tech University VORTEX2 Team
While others seek shelter from the storm, Texas Tech researchers go face to face with Mother Nature. These brave scientists are trying to solve the mystery of tornado genesis, and after six weeks of intense storm chasing and data collection, they just might have some of the answers.
Researchers are now beginning to dig through the data gathered during the recent Verification of the Origins of Rotation in Tornadoes Experiment 2 (VORTEX2), an $11.9 million project funded by the National Science Foundation (NSF) and the National Oceanic and Atmospheric Administration (NOAA). The 2010 field portion of the project lasted 45 days and involved about 120 people, including 18 Texas Tech researchers, more than 40 vehicles and an array of data-collecting instruments from Doppler radars and unmanned drone aircraft, to StickNet probes and weather balloons.
The purpose – to discover the origins of tornadoes in order to better predict when and where storms will pop up, determine how severe they will be, and in turn, improve alert systems by increasing warning times.

Read the Article


Friday, June 11, 2010

A classic, isolated supercell....finally

Hello out there! If you've been keeping up with us on Facebook and Twitter, you've seen that we've been pretty busy the last week or so. We've covered every mile of I-80, two or three times for some stretches. Last night we finally found ourselves operating back in Colorado. Storms were initiating just off the mountains, and found themselves in favorable environments, so we spent the afternoon scooting down there from northwest Nebraska, where we had stayed the previous night.

Yesterday actually brought two targets, and therefore two deployments for us. The first supercell had a rather impressive wall cloud for quite a while. StickNet team 3 managed to get a few probes down before it became apparent that a storm coming up from behind was going to overtake the target storm. We therefore shifted our attention to the southern storm, which turned out the be the right decision!

My team was the eastern most team of all StickNet teams, so we had a great view of the storm as it approached from the west. We saw the first tornado touch down from 35 miles away! I would say it was on the ground for about 5 minutes before it either dissipated, or became to rain-wrapped for me to see. So I'm sure you're asking, "where's the tornado picture?" Well, I was driving, trying to get to our deployment road, so no picture-taking while driving for me. The mobile mesonet and radar teams reported a second tornado not long afterward, but our view was too obscured by rain to see anything.
My team made deployments on two roads, US 36 east of Last Chance, CO, and Colorado Highway 63 north of Anton. We placed four probes on the first road, then took off to the east to make the northward turn out of Anton. I had Chris Weiss on the phone with me for the last two deployments, as he was trying to position us exactly so that the mesocyclone of the storm would pass right over those last two probes. Its not often that he says, "well, where I'm going to place you depends on how soon you make the north turn." This is more what you would expect for the fine-scale array teams (i.e. non-trailer teams). But on this particular day, because of the huge gap in roads, and the spread between our StickNet teams, my team actually was working on a fine-scale array, needing more precision than our typical course array.

After making the last two probe drops, we headed south for safety near Anton, and watched the storm for a while. The picture I'm including was taken at this point, after we'd finished the deployments and were waiting for the storm to pass Highway 63. This storm was, by far, the most photogenic storm I've seen this year. All the features were very clear and well-defined. Storms like this make our jobs much easier, as we know exactly what we're looking at, and can adjust our positions and the arrays accordingly.

Well, I guess that storm decided it did NOT want to be sampled, because it made a huge looping turn to the north and then northwest as it approached those last two probes. It never crossed them, as the storm began cycling just before Highway 63. What a disappointment! Oh well, there's nothing we could do about it. We made the best deployments possible given the road network, and the previous motion of the storm. The storm did pass over our array on US 36, and through the arrays laid out by the other three StickNet teams. And the other V2 teams were hard at work collecting data as well. From what I saw during operations last night, it looks like the radar teams did a good job of covering every phase of the storm.

So far, we are unsure of any damage reported with this storm. It mostly passed over open land. It did pass near the town of Last Chance, and one of the other StickNet teams reported that they saw some mobile homes that may have had some roof damage, but were unsure if this was caused by the storm, because of the lack of damage to nearby trees and power lines. I will be driving through there in the next hour, so I'll have a look myself. It may have been caused by last night's storm, or could be pre-existing damage. I hope I am able to tell the difference. If the damage was caused by last night's storm, knowing the wind speeds in the area would certainly be valuable.

Well, back to the driving for me.....

Wednesday, June 2, 2010

Rolling with the Punches

Good morning loyal followers! I know it has been quite some time since I have blogged but an early departure has allowed me to get a jump start to my day, affording me plenty of time to fill all of you in with what has transpired lately.

After much debate a major change to operations has taken place. The leaders of VORTEX2 have decided that earlier departures and briefings on the road would grant the project more time to reach distant targets that would otherwise be unreachable based on prior departure times. Basically, we are transitioning into a pseudo desperation mode with two weeks left to go! We have yet to accomplish some of our project goals including intercepting a vigorous tornado and we need to pursue every opportunity that Mother Natures serves up to us.

Intercepting tornadoes is no easy task as we still do not have a complete understanding of the environment that they form in. A perfect example is the Memorial Day tornado of this year that occurred in southeast Colorado. The setup was marginal and ultimately the final decision to position for the next day's weather event was made. An eventual tornado formed out of the marginal conditions and another missed opportunity was at our hands. Attention had been given to the Colorado target area in the morning briefings but the conditions were not favorable enough to justify heading south to Colorado. This just goes to show how rigorous our science is and how challenging point forecasting can be.

This transition that we are making now will indefinitely put more stress on all teams as the days will get longer and the nights shorter, especially when we start deploying our instruments. The nights will get even shorter as instruments will need repairs (hopefully not)!

Despite the increased level of stress, our team is remaining focused on collecting as much data as we can and representing Texas Tech University proudly. These are times where we all have to put pride and short tempers to the side and remain patient. It is very easy to let loose and lose your cool, but I must say our team has done a fine job keeping our composure. We pride ourselves on utilizing our resources efficiently in the field and know that by exercising patience, Mother Nature will surely reward us with some adequate data sets.

Stay tuned for more exciting news from VORTEX2 Land!

Tuesday, June 1, 2010

No one ever said this was easy.....

Hello out there in blogger land! The last week or so has been weird for the V2 team. And has proved just how much we still need to learn. And has shown just how difficult this project can be.

The federal government (and you as taxpayers!) have charged us with finding out everything we can about tornadoes, how and why they form. The problem is in the logistics. These days, with so many chasers, and the abundance of media, and cell phones, and cell internet, seeing a tornado is not as hard as it used to be. But making meaningful deployments of instruments to collect valuable data to understand tornadoes is incredibly difficult. Because V2 is so huge, we need a long lead time (~1 hour), and a good road network (we're trying to get 50 vehicles to sample a single storm) in order to carry out our objectives. And in the Plains, where the weather is volatile, this is a daunting task.

We've been at this for 31 days, and have collected a few great datasets. But we've also collected a bunch of mediocre ones as well, where storms were dissipating as they crossed the instruments, or made an unexpected turn, or didn't make the turn we thought they would, and just skirted the edge of the arrays. There's also the cases where the storms were really good (when I say good, a non-meteorologist would say bad), but moving so quickly and were so violent that we could not get a full array of instruments in position without sacrificing our own safety.

And at some point, we all reach a breaking point, just physically and mentally exhausted. 31 days of sharing hotels, driving late at night, staying up to fix instruments, eating fast food or missing lunch or dinner altogether, visiting the laundromat in our "free" time. On a few rare occasions, the project leaders have decided to give us a down day, to let us all regroup, and remedy any instrument issues. We have missed a few events because of this, but I think they were keeping our well-being in mind. We can't always drive 5 hours to operate, operate for 5 hours, then drive another 6 to get back in position for the next day. We can do it occasionally, but not multiple days in a row, we just run out of awake and alert people to drive. Missing events leads to some frustrated people. But then again, 4 hours of sleep 4 days in a row also leads to some grumpy people, so there's always a trade-off! After no real operations for the last several days, I am hopeful that we're all rested and prepared for these last 2 weeks. With only 2 weeks left, I think the crews are ready to just push through til the end.

Here's hoping that yesterday's decision to forego operations to give us a better chance (position) for today will work out for one ever said this was easy. If it was easy, everyone would do it.

Monday, May 31, 2010


Tough day today. We chose to pass on the Southeastern Colorado target for better day 2 positioning. As it turned out, the southern target did produce a few tornadoes, though we have no idea how targettable they were (i.e., how well they would have satisfied the low-level wind field objective of VORTEX2, the only area of our core objectives that has remained largely unsatisfied thus far in the project). Though frustrating, the day well underscores how much we have yet to learn about tornado forecasting - why we have this project in the first place!

Further, today illustrates the myriad of choices that the investigators on this project have to face on a daily basis. As a coordinator of two observational platforms and ~20 participants, I face a number of choices over the course of a typical day. Some of these are relatively straightforward, stopping for gas or lunch, for instance. Others are more involved.

One of the more difficult choices I face is the decision to deploy StickNet probes. Though we have on occasion had multiple deployments of the same StickNet probes in a given day, we usually cannot count on that luxury. Therefore, the decision to start a deployment is not taken lightly at all. Many factors have to be considered, including the number of probes remaining, the available road network, the current quality of the target, the anticipated future of the target, the development of later targets, the amount of daylight left, the time available to make a quality deployment, hail distribution, lightning frequency, and so on. The spacing, orientation and width of deployments has to be decided on the fly.

Occasionally it is necessary to abort StickNet deployments due to safety concerns. We had such a situation back on 10 May in Central Oklahoma. A dangerous tornadic supercell thunderstorm was bearing down on the lead deployment team. [Owing to various factors, we got a late start on the lead array]. The decision had to be made to bail on the deployment. We never sacrifice safety at the cost of a dataset, no matter how good it might potentially be.

Other choices revolve around staffing, both in deployment and lodging assignments. Crew members get sick or injured, family emergencies arise, students need to get experience in new tasks, personalities conflict (on rare occasion). All of these things have to be addressed to ensure an efficient (and happy!) team. Thankfully, the TTU team works together pretty smoothly, which makes life much easier.

Back to the forecast choices made today - we will see if the decision to prioritize day 2 was the proper one. Even in the best of seasons, mistakes will be made that are only clear in hindcast. The frustrating aspects of these misses are also motivating though - we still have plenty to learn!

Sunday, May 23, 2010

Can You Find the Tornado?

A big hello to all 25 of our followers and anyone else who may be reading this! I'm always a bit unsure of what you all may be interested in reading about from VORTEX2, so I'm going to shamelessly attempt to run up my comment total by asking anyone with specific questions or suggestions for posts to let me know and I'll try my best to answer them in subsequent postings.

In the meantime, I will just share a few images the TTUKa radars have captured in the past week. The above image does have a tornado in it, which occurred to the northwest of Kingfisher, OK on May 19th, but it's not easy to tell at first glance. The image is a range-height indicator scan (which is a vertical slice through the storm) with reflectivity on the left and radial velocity on the right. The tornado is the area of lower reflectivities (green colors) near the lower center of the left panel. For this image, we were scanning the westward edge of the tornado from the south, so the winds are blowing towards the radar (cooler colors in the velocity image). The reason for the large area of orange colors within the tornado in the velocity image is that the wind speed is too high for the radar to unambiguously sample. So, once the velocity passes a certain wind speed it "folds" to the opposite end of the color scale, like Pac-Man wrapping around to the opposite side of the screen. So, backing out the actual velocity, we estimate wind speeds of upwards of 50 m/s or just over 100 mph, and we likely didn't sample the highest wind speeds in this tornado!

Ok, the second one is a bit easier, especially if you've read my last post. This image is a near-surface PPI (horizontal plane) scan taken by TTUKa-2 of a weak tornado east of Dumas, Texas on May 18th. One of the cool things about this tornado is that it did not occur in the mesocyclone, but along the rear-flank downdraft boundary (RFDB). This storm exhibited many strong circulations along the RFDB that we could watch wrap back into the mesocyclone. As circulations like this have been theorized to contribute to tornadogenesis, capturing their evolution with our radars is a very exciting finding.

Finally, this third image is not of a tornado, ... I think. This is a close-up image of the very tip of the hook of the Dumas supercell. So, when you look at a hook echo from a standard radar, the big red reflectivity ball in this image would be roughly representative of the trailing one or two pixels. All of the tip of the hook in this image is within five kilometers of the radar and TTUKa-2 was set up about 3 miles to the east, scanning the same mesocyclone. This means that we can conduct a dual-Doppler synthesis of this event and derive the full three-dimensional wind field of the near-surface mesocyclone! (this is the kind of nerdy sentence we meteorologists like to put exclamation points at the end of). So some lucky graduate student (possibly me I guess) will get to spend countless hours sorting through the fine-scale details of this case trying to make sense of it. Woohoo!

Well, we're headed to bed tonight in a state that starts with the letter "K" and should be operating again tomorrow, most likely on a road somewhere. I do want to mention that I've run the bird total up to nearly 50. Highlights include three life birds for me, a Cassin's Sparrow and Scott's Oriole just north of Carlsbad, NM, and several Clarke's Grebes in North Platte, NE yesterday.

Finally, finally, one last image, of the Kelvin-Helmholtz instability waves propagating vertically I mentioned in my last posting:

Thursday, May 20, 2010

The Dumas Day.....What a day!!

So I'm definitely a bit behind on blogging, but this day truly deserves its recognition. On Tuesday, May 18, our deployment lead us to Dumas, TX, just northwest of Amarillo, TX. This ended up being a very impressive supercell and came dangerously close to dropping a tornado over the town during rush hour.......not good at all. Luckily, only a funnel cloud was reported as it made it's monstrous treck across the middle of town, although others were reported east of town. Our deployment was relatively uneventful at first in that we didn't really have any challenges with terrain or power lines, but traffic was somewhat of an issue, given the time of day and the fact that it is a fairly major road. We did, however, struggle with the placement of our last probe as we had approached town. We aren't really able to deploy in areas like that as it presents the danger of flying debris damaging the probes, should a town be unfortunately struck. We also are not able to get the proper "exposure" to the instrument. What this means is that if there are, say, a lot of trees near a deployment, the wind speed and direction can be greatly affected by those obstacles and can skew the data actually recorded by the probe. The wind could be enhanced or hindered and thus we would not get a true sample of what actually happend within the storm as it crossed the array. This, among other things, is why we look for wide, open areas in which to stake our probes.

As for the storm, it had a very impressive structure as it crossed the array, with a large hook echo headed straight for Dumas. Large hail had already been reported from this storm at this point in time, not to mention the fact that it could have produced a tornado at any point in time as it approached Dumas. It was somewhat disturbing that during our deployment, the tornado sirens within Dumas were NOT sounding off, which worries me a bit. If the storm had dropped a tornado, it could have been even worse given the rush hour traffic and many people could have possibly been left without enough time to take adequate shelter. As we finally reached the south side of town, we found a good deployment location and dropped our last probe. Once we were finished, we basically sat there, watching and waiting to see what would happen. Luckily, no tornado was produced within town, but either way, the town was rocked with high winds and hail. As the supercell finally passed, we were able to make our way back to the north to our first probe. Thankfully, we saw no damage to anything within town, but we reached a point just north of the center of Dumas where marble sized hail covered the roadways. You could even see fog covering the fields as the hail was sublimating, or turning from a solid directly into a gas. Our first probe was unfortunately separated from us by a huge ditch full of icy hail-water......which is just as much fun as it sounds. One of my team members was able to capture a few candid shots of the pick seen above, so enjoy! The probe deployed closest to the north side of Dumas still had golfball sized hail stones and the nose cone on the propeller anemometer was broken off, not surprising given the hail. Some of the other StickNet teams were slightly more unfortunate as they encountered baseball sized hail. Let's just say that day lead to a pretty late night of fixing damaged connectors and replacing pieces of instrumentation on the probes. That's what baseball and tennis ball hail can do if caught outside.

Overall, this day was probably one of the most successful deployments so far for all of the VORTEX2 teams. Everyone was able to sample the storm from start to finish and collect very good and very valuable data. Our specific deployments were textbook and our team definitely could not have done a better job that day. It was a proud day for all researchers involved and hopefully we will be as successful in the future for the 2010 season.

Well, I guess that's all for now. After a full day of traveling, I'm kind tired ;) Until the next one....