March 2020 Tornado Outlook

Predicting severe weather, specifically tornadoes, on a monthly or seasonal basis is a very difficult task. The variability in smaller scale weather patterns that affect tornado potential is large, and grows significantly as you move further out in time. This means that medium and long range tornado forecasts can naturally tend to be quite ambiguous.

With that being said, the medium range weather forecasting process in itself can offer us some clues into the weather pattern that will hopefully provide a glimpse into the tornado potential that lies ahead. On a monthly lead time, it is possible to analyze subseasonal forcing and teleconnections, the ongoing hemispheric waveguide, and analog years to create a picture or an identity for the month ahead.

The Winter of 2019-2020 has featured some very specific and notable features, but none more anomalous than the highly positive Arctic Oscillation, or AO, which reached near all-time records twice in February of 2020. The only other time this occurred in February was in 1990 (which also happens to be a reasonably good analog for the Spring ahead, but we’ll talk about that later). Concurrently, the stratospheric polar vortex was one of the strongest on record, exceeding ERA maximum values more than once during the DJF trimonthly period.

This is an important characteristic of the Winter, as it indicates a strong polar vortex and arctic jet stream in the high latitudes. This affects the pattern in the mid latitudes, too (think: lack of high latitude blocking and effects on baroclinicty). I built my analog composite for the Spring ahead using data from the high latitudes including the NAO and AO – and quite frankly there were not many years that fit the bill. This year was very anomalous in almost all respects when it comes to the higher latitude weather pattern.

The pattern in the Pacific Ocean is also of importance – and there were a few very important features that re-appeared several times throughout the Winter. Perhaps most importantly is the January and February data suggesting a weak negative PDO, or Pacific Decadal Oscillation. The PDO for January and February 2020 was -1.40 and -1.35 respectively, their lowest values since 2013. You can read more about the PDO and its impacts on the weather pattern here.

Obviously, what’s occurring in other parts of the globe remains highly important as well – including but not limited to ENSO conditions (weak El Nino), the AAM/GWO, current impacts of the MJO (including the recent weakening of a major +IOD anomaly). Factoring in all of these ongoing conditions – and their state over the past few months – it is possible to filter through past data and create a list of analog years that behaved at least somewhat similarly. While the playing field was thin in terms of analogs this year, I was able to create a weighted composite that I felt comfortable with.

One thing stood out about the analog years: Many of them featured active severe weather/tornado seasons, and many of them featured an active March in terms of tornadoes.

Looking at the upcoming pattern on global ensembles, I was struck at how similar the 7, 15 and 30 day height anomaly composites at 500hPa were to past March severe weather outbreaks. When I created a composite of March’s featuring severe weather outbreaks in the past 50 years, the composite matched almost perfectly with what is being suggested by a large majority of medium and long range guidance.

With all of this in mind, my March 2020 forecast features a heightened chance of above normal tornado activity across the United States. In particular, it appears likely that tornado activity could average above normal from the Arklatex into Dixie Alley, although some events may also occur in the Plains during mid to late March. I’ve broken the month down into the following percentages:

Chance of above normal tornado activity in March: 50%

Chance of normal tornado activity in March: 30%

Chance of below normal tornado activity in March: 20%

As you are probably aware, this does not guarantee that a tornado outbreak will occur. This is simply a forecast – largely speculative – based on my knowledge and meteorological opinion.

If you live in areas highlighted as potentially above normal for tornado activity in March, do not panic. Simply have a plan in place in case a severe weather event does unfolds, as you always should.

I’ll revisit this post in late March to see how I performed in hopes that I am able to continue building on monthly tornado forecast skills and learning along the way.

Cheers until then!

John