Feb 15th 2017 3:15 am
Disclaimer: My incoherent ramblings are always based on as much fact as I can find. If any experts happen across this and I’ve said something incorrectly please do tell me. I’m always open to learning something new. I only hope I can pass on a little knowledge to help make the world a bit better.
Spring is just around the corner, and I mean just around the corner. March 1st marks the beginning of meteorlogical spring, and more often than not, the beginning of the traditional storm chase season. That date also brings with it numerous rituals of cleaning out the cars, dusting off the cameras, and preparing for a 3 month race to get the best video of the year. But before we get too excited, what does tornado season actually mean? This image from USTornadoes.com really send the message home. March usually marks the first significant uptick in tornadic activity after the winter lul.
But as many know, tornadoes can happen anytime of the year. We have already seen a deadly start to the season with 138 preliminary reports in January and 20 recorded fatalities. The data, courtesy of SPC shows that January averages around 16 reports and 0 fatalities. Nearly 9 times less than what we have seen already. However 3 out of the last 4 years have been below average years in the number of tornadoes.
ZCZC STAMTS ALL NWUS21 KWNS 280543 TORNADO TOTALS AND RELATED DEATHS...THROUGH FRI JAN 27 2017 NWS STORM PREDICTION CENTER NORMAN OK 1143 PM CST FRI JAN 27 2017 ...NUMBER OF TORNADOES... NUMBER OF KILLER TORNADO DEATHS TORNADOES ..2017.. 2016 2015 2014 3YR 3YR 3YR PREL ACT ACT ACT ACT AV 17 16 15 14 AV 17 16 15 14 AV --- ---- ---- --- ---- ---- ---- -- -- -- -- -- -- -- -- -- -- JAN 138 - 18 28 4 16 20 2 0 0 0 3 1 0 0 0 FEB - - 102 3 42 39 - 7 0 0 2 - 4 0 0 1 MAR - - 85 11 20 38 - 0 1 0 0 - 0 1 0 0 APR - - 140 171 129 146 - 1 2 35 12 - 1 1 8 3 MAY - - 216 381 130 242 - 2 7 0 3 - 2 5 0 2 JUN - - 86 184 286 185 - 0 0 2 0 - 0 0 2 0 JUL - - 107 115 85 102 - 0 0 4 1 - 0 0 1 0 AUG - - 89 45 33 55 - 0 0 0 0 - 0 0 0 0 SEP - - 36* 17 41 31 - 0 0 0 0 - 0 0 0 0 OCT - - 23* 40 73 45 - 0 0 1 0 - 0 0 1 0 NOV - - 69* 99 23 63 - 5 0 0 1 - 2 0 0 0 DEC - - 14* 83 20 39 - 0 26 5 10 - 0 6 2 2 --- ---- ---- ---- ---- ---- ---- -- -- -- -- -- -- -- -- -- -- SUM 138 - 985 1177 886 1041 20 17 36 47 29 3 10 13 14 8 *PRELIMINARY REPORTS. PREL = 2017 PRELIMINARY COUNT FROM ALL NWS LOCAL STORM REPORTS. ACT = ACTUAL TORNADO COUNT BASED ON NWS STORM DATA SUBMISSIONS.
Early season numbers don’t always paint the picture for the rest of the season. Tornado season is fluid. It encompasses a broad range of climatological and synoptic patterns that make nailing down any distinguishing characteristics very difficult from a far. Despite the complex nature, some research has shown weak correlations to well established climatological patterns such as ENSO. Much of the information was gathered from a presentation by:
Robert Atlas1, David Enfield2,1, Chunzai Wang1, and Hailong Liu2,1 1NOAA-AOML, 2Univ. of MIAMI-CIMAS
Is there an optimal ENSO pattern that enhances large-scale atmospheric processes conducive to tornado outbreaks in the U.S?
Their research focused on correlations between ENSO phases and major US tornado outbreaks. They also discussed the use of the Trans Nino Index (TNI), a Normalized SST difference between Niño 1,2 and Niño 4 regions. Supposedly this captures the initiation and decay phases of ENSO better than traditional means. The beginning and ending phases were found to have some weak correlations with tornado activity across the US. A quick summary of the presentation follows. I highly recommend giving the presentation a view as it has some nice stats showing the fickle nature of correlations.
“Summary and Discussions
- ● Observations and reanalysis products are used to show that a positive phase of the Trans-Niño is linked to U.S. tornado outbreaks.
- ● The TNI-U.S. tornado link is due to the enhanced large-scale differential advection during a positive phase of the Trans-Niño 1) anomalous upper-level cyclone over the North America;
2) increased GoM-to-U.S. moisture transport;
3) increased lower-level vertical wind shear east of the Rockies.
- ● Lee et al. [2013, JCL in-press] used model experiments to explore the potential mechanisms for the link between Trans-Niño and U.S. tornado activity.
● Positive Trans-Niño frequently occurs during either the onset phase of El Niño or the decay phase of La Niña (not shown).”
In essence, positive TNI correlates the best of any parameter to large scale US tornado outbreaks. Large positive TNI anomalies are found often with weak La Niña trending towards El Niño, or El Niño decaying. (Note more research may be needed on this.) Their research concluded that TNI had a low probability to predict major tornado events across the mainland US. We may also be able to use this (very very cautiously) to try and glean some information about the upcoming year.
Current CPC information had a strong El Niño in 2016 that decayed to weak La Niña and to ENSO neutral by early this year. Gif of the SST anomalies gives an idea of the change over the preceding months. Notice the lack of any really strong warm or cold tongue. This is classic ENSO neutral.
CPC also does a weekly ENSO forecast and have detailed some interesting findings. ENSO neutral is expected to continue into spring of 2017 with a slight chance of a weak El Niño developing. While up to date TNI data is not available, the latest data indicated a trend towards positive values. Given the change over to El Niño is possible I would expect this positive trend to continue. If what I mentioned earlier holds true that COULD mean a potentially more widespread or active tornado season. However these correlations are weak and subject to many many variables it is difficult to gain a true idea of what will happen.
One other tool we can use to look at are the CPC monthly forecasts. They can be somewhat skillful in nailing down large scale patterns. For instance the 6-10 day forecast shown below calls for above average temperatures across the eastern US with below average to the west.
This would likely indicate upper level ridging across the eastern US and troughing along the west coast. This pattern is well known for being an active storm pattern with strong cyclones passing over the CONUS. The 3 month outlook is much less telling, not indicating much, other than a possible northern shift of zonal westerlies.
Other tools such as the CFS can be of use as well. Greg Carbin former WCM of SPC designed a cool tool that shows long range forecast potential from the CFS. Often referred to as the “chiclet chart”, its become one of my favorite tools for recognizing potential up to 45 days out. The chart shows repeated CFS forecasts for supercell composite and QPF and then tallies the number of grid points with SCP >1. This gives a rough estimation of supercell environment over a long time range. The best part is the ability to see run to run consistency by matching vertical columns with brighter colors. This indicated the model has “locked on” to a system and has consistent timing. While its difficult to really gather specifics, the tool is a nice “heads up” that something could be occurring. There is also a seasonal component noticeable. As the spring season rolls around conditions more favorable for storms begin to emerge. This manifests itself as noise (random increase in gridpoint counts) on the far right of the image. While a clear signal is not present, the general increase in noise could indicate an upturn in convective activity. Given its timing associated with the beginning of spring, thats not too much of a surprise. Overall the tool gives a nice look for long range potential.
In summary, theres not a great signal for this year. Its messy and somewhat conflicted. ENSO is neutral with only a slight preference towards EL Nino, the trend over the last few years has been below average numbers or tornadoes, TNI is weakly positive but continuing to trend up, and CPC data indicated potential for a more active short range with a less predictable 3 month time frame. If I had to make a prediction (If you can call it that?) my best guess would be that we will continue to see an active early season over the next 1-3 weeks. West coast troughing should continue to keep an active storm track over the CONUS. This may keep things active in the short term, but the long term may see only an average to slightly below average year, with more activity east of the Mississippi river than the past few. This keeps inline with the more broad geographic outbreaks associated with positive TNI but relatively weak values on tap so far this year. Just for fun with absolutely no scientific reasoning, ill estimate the number of tornadoes and some other stats by year end. I hope this was fun and somewhat useful. Its likely horrendously wrong, but thats okay! Maybe ill get better if I keep doing these!
2017 tornado estimate: 1,005
Outbreak days (10+ tornadoes): 8
Most active week: May 21st-28th
Highest Rated Tornado: EF-4
First Plains Chase day: March 2nd 2017
Number of tornadoes I see: 21