Saturday, January 1, 2011

Reviewing yesterday's outbreak and comparing it to January 7-EIGHTH, 2008

In my last post I compared a GFS forecast for this Friday's severe weather event to what actually happened on January 7th, 2008--another wintertime severe weather event when we had a strong tornado move through northern Illinois (where I live) and then several other tornadoes further south in Arkansas and Missouri. I had concluded based on the forecast that things didn't look as good for a similar severe weather event despite superficial similarities in the forecast surface synoptic setup.

Well, that was only partially right.

Indeed, northern Illinois (where I grew up and was focusing my efforts) did NOT see another strong tornado, or really any severe weather at all--just some rumbles of thunder as the front came through in in the areas of heavy rain that proceeded it.  But look at the storm reports from yesterday:
Fig 1 -- SPC storm reports for Dec. 31, 2010.
Particularly, compare this to the storm reports for January 7th, 2010.
Fig 2--SPC storm reports for January 7th, 2010.
There is a roughly similar area of severe reports in central and southern Missouri and into northwestern Arkansas.  Sadly, at least six people died in the storms that moved through yesterday through these areas and there was significant damage to several towns.  However there was no severe weather in northern Illinois or southern Wisconsin yesterday and there was also a lot of severe weather in central Mississippi yesterday.  So there are distinct differences.  Then I just happened to click one day forward in the SPC storm reports to see what happened on January 8th, 2010:
Fig 3 -- SPC storm reports for January 8th, 2010.
Here we still see storm reports in the northern Arkansas region, but we also see much more activity in Mississippi.  So this event from yesterday seemed to combine factors from both of these days.  Want to see something really eerie, though?  Compare yesterday's 12Z surface analysis from the HPC, particularly noting the location of the cold front from northern Illinois down to the Texas border:
Fig 4 -- HPC surface objective analysis from 12Z, Dec. 31, 2010.
To the same analysis from 12Z on January 8th, 2010:
Fig 5 -- HPC Surface analysis for 12Z, Jan 8, 2010.
The fronts are almost in the exact same position, with a surface low analyzed at both times in northeastern Oklahoma/southeastern Kansas.  No wonder we're seeing interesting similarities in the locations of the severe weather events.  Furthermore, look at the temperatures and dewpoints from observations on those maps.  For example--at Little Rock, Arkansas--on Dec. 31? 66 degrees with a 62 degree dewpoint.  On Jan. 8, 2008?  68 degrees with a 61 degree dewpoint.  At Jackson, MS--Dec. 31? 66 degrees with a 61 degree dewpoint.  Jan. 8, 2010?  65 degrees with a 60 degree dewpoint.  Very similar in terms of warmth and moisture.  Let's look at the surface analyses for 00Z--twelve hours later.
Fig 6 -- HPC surface objective analysis for 00Z, Jan. 1, 2011.
And for the January 8th event:
Fig 7 -- HPC objective analysis for 00Z, Jan. 9, 2008.
There are immediate differences noticeable, particularly in terms of the pressure field.  In the Jan. 8th event, there are much, much stronger pressure gradients and a very dominant low pressure center over northern Michigan.  By contrast, yesterday there were much weaker pressure gradients, and there was an occluded front (a sign of a filling low) accompanying the surface low back in central Minnesota.  But the position of the cold front from the Memphis area down south to the Gulf Coast is rather similar (it's slightly further east on Jan. 9th, but still--roughly similar).  So we have a cold front moving through very similar air masses in Mississippi and Arkansas at about the same time of day in both events.  This helps to explain why the Mississippi storm reports were present on both days.

So we can reasonably see why there was the similarity in Mississippi on both days.  So why didn't we see reports in Missouri on both days?  My guess is in the differences in the movement and orientation of the more northerly half of the system.  First, in figure 5 above, we see that the "cold" front from northern Illinois down through central Missouri is marked as a stationary front at 12Z on Jan. 8th.  In fact, this frontal boundary was around since the day before:
Fig 8 -- HPC Surface objective analysis for 00Z, Jan. 8, 2008.
We can see that there still is a cold frontal boundary from northern Illinois through central Missouri.  But note how far the cold frontal boundary south of that point had to swing in 24 hours to be at the Mississippi River like we see in figure 7 above.  That's a true cold front--the cold air is definitely advancing.  But with the frontal boundary already in central Missouri on the 7th, this provided a forcing mechanism to help generate convection there a day earlier.  Since the primary function of convection in the atmosphere is to restore equilibrium (and stability), by the time the front finally pushed through on the 8th, there's a good chance that the atmosphere had already stabilized some.  Let's check the central Illinois sounding for 00Z, Jan 8th as convection was going on:
Fig 9 -- KILX sounding from 00Z, Jan 8, 2008.
Note the respectable surface CAPE values of 529 J/kg and some rather steep lapse rates in the mid levels, indicating good instability.  We're also rather saturated with a nice backing of winds from the surface up through 700mb--signs of warm air advection and good wind shear for rotating storms.  Unfortunately the 12Z sounding from the next morning is not complete:
Fig 10 -- Incomplete sounding from KILX at 12Z, Jan 8th, 2008.
This isn't even an official sounding since the balloon only got up to 700 mb.  But, if we were to assume it is at least somewhat correct, we can see a pronounced isothermal layer up to around 800 mb.  The wind profile still has southerly winds at the surface, so we know the cold front has not passed through yet.  However, our instability seem to have been greatly reduced.  Of course, we suspect this sounding is bad--note that the lapse rate looks like it is superadiabatic (steeper than adiabatic) right around 800 mb.  This would be absolutely unstable and immediately convect if it were actually that steep of a lapse rate.  Since we practically never see lapse rates that steep in nature (because it would immediately convect), we suspect that this sounding has messed up in some way.

Still, my point is that convection on January 7th in the Saint Louis area probably helped stabilize the atmosphere enough to limit any further convection in that area on January 8th.  However, in our event yesterday, there was no convection the previous day and as such everything happened at the same time.

I wanted to compare upper air charts as well (since they are remarkably different) for these two events, but this is already a long post and apparently Blogger has an image limit of 10 per post that I was not aware of.  So, I'll do another quick post tomorrow to finish this up and compare the upper air patterns.  For a brief preview of that, though the depth and placement of the troughs are vastly different, the upper-air winds over our areas of interest are very similar.  So in terms of shear and any other sort of forcing due to jet streak divergence aloft and whatnot, things are actually rather similar, even though the overall patterns themselves are rather different.

But anyhow, that was a look at how two wintertime severe weather outbreaks were actually very similar in terms of their structure and timing (at least at the surface).  I want to especially thank Tim Supinie for suggesting that I look into this.  Since there was no severe weather yesterday in northern Illinois I had originally not thought to revisit the comparison, but this look back does highlight some nice similarities between the two events.

No comments:

Post a Comment