Saturday, April 2, 2011

Thoughts on Severe Weather Sunday Night, Part 2

Now that I'm back again, to continue the discussion.  I already noted in part one how the upper-air pattern seemed somewhat chaotic in the forecast and that would also reflect on the surface pattern.  It's the surface and low-level variables that really have people worked up about this event.  Remember how we saw a strengthening pressure gradient due to pressure rises and falls across the middle part of the country?  That means increased southerly winds across the plains and the midwest.  With those winds comes strong advection of moisture and warm air northward.  Here's the NAM model forecast for dewpoint temperatures tomorrow evening:
Fig 1 -- NAM 36 hour forecast of dewpoint temperatures (colors) and winds (barbs) for 00Z, Monday (Sunday night), April 4, 2011.  From the HOOT website.
Note that the 60 degree dewpoint contour is all the way up to northern Illinois and eastern Iowa by Monday evening--that's very far north to be seeing dewpoints that high.  Particularly when the 70 degree dewpoints (the darker greens) are still confined to the Gulf of Mexico.  The southerly winds are very evident within that tongue of moist air moving north.  However, the wind pattern and the shape of that warm, moist sector are somewhat odd on the western side.  Note we have several westward "bumps" in the moisture coincident with strong wind shifts.  These are marking areas of perceived low pressure and surface boundaries.  I personally would interpret the above pattern like this in terms of surface features:
Fig 2-- Same as figure 1, but annotated with surface features by the author.
Each one of those "bumps" westward has somewhat of cyclonic flow around it and a quick check of the surface map (not shown) shows both areas to have relatively lower pressure than their surroundings.  So, both are labeled as "lows" on the map.  The leading edge of the moisture surging northward would be the warm front--that's pretty straightforward.  Strong north-westerlies behind the northern low look like flow behind a cold front, so I've drawn a cold front along that boundary.  Now, what's happening around the southern low is interesting.  Usually when we see a shift in the winds with north to northwesterly winds to the north and more westerly or southwesterly winds to the south, the wind shift to the south represents a dryline and the winds to the north show a cold front.  This wind shift also seems to extend westward behind the southern low.  I've labeled it with a dotted line above, indicating some kind of surface trough.  However, all I believe has happened here is that the cold front associated with the northern low is overtaking the dryline to the south.  Eventually those northwesterly winds behind the cold front will sweep in behind and there will be no more dryline.  The places where these boundaries intersect are known as "triple points", and you may see them mentioned in weather discussions.  That southern "low" would be considered a "triple point" since it's an area where three boundaries meet--the cold front, the dryline and the surface trough.  Areas near triple points in the moist region (so, to the east in the above map) are often favored areas for severe weather development.

So we see a complex suite of boundaries and low pressures in the forecast--reflective of the complex upper-air pattern we saw above.  The boundaries (and to an extent the upper-level winds) will provide the lifting mechanisms, and we can see in the dewpoint maps above that there is moisture available to work with.  We need two more things for severe storms--wind shear and instability.  With respect to wind shear, there is an awful lot of that.  The 850mb winds are forecast to be roaring out of the southwest on Sunday evening.
Fig 3 -- NAM 36 hour forecast of 850mb winds (colors) and geopotential height (contours) for 00Z, Monday (Sunday night), April 4, 2011.  From the HOOT website.
Those are 65 knot+ winds in the Ohio River Valley on Sunday evening.  Note that the wind swath seems a bit east of the fronts and the dryline, but there are still strong winds over the entire moist sector.  Even though the strongest winds may not be exactly over the region with the best forcing, there still are strong enough winds just off the surface for there to be plenty of low-level wind shear.

What about instability?  A good first glance at instability is to look at the surface-based convective available potential energy (CAPE).  The higher the CAPE, the more the instability.
Fig 4 -- GFS 36 hour forecast of surface-based CAPE for  00Z, Monday (Sunday night), April 4, 2011.  From the HOOT website.
Strong CAPE values are pooled near the dryline and cold front, but also extend out east (somewhat more weakly) into the warm sector.  But, just because there's CAPE doesn't mean that there will be storms.  That available convective energy is only "available"--any storms have to be able to tap into that.  With a strong lifting mechanism (like a cold front, for example...) it becomes very easy to tap into that energy because the lifting is being done by the advancing cold air behind the cold front.  As such, storms are much more certain to fire along the cold front up north as opposed to further south, even if the CAPE is higher to the south.

So what would be stopping storms from forming and becoming severe further south?  The answer is something called the "capping inversion".   Normally air cools with height--it's much colder further up in the atmosphere than it is near the surface.  In fact, the faster it cools off with height, the more unstable the air gets because warm air near the surface is more buoyant than the colder air above.  This promotes rising motion and the formation of thunderstorms.  In fact, in some ways, CAPE is a measure of how steeply the temperature cools with height (it includes other things like moisture content too...).  Generally, the faster the air cools with height, the more CAPE you have and the more unstable the air becomes.  The opposite happens in a "capping inversion".  This is a region of the atmosphere where the temperature actually warms with height.  This is bad for rising air because if it runs into air that is warmer than it is, the rising air is no longer more buoyant than the air around it and it will stop rising.  Capping inversions (or, "the cap" for short) are easy to see on vertical soundings.  Here's a NAM model forecast sounding for Norman, Oklahoma at 21Z on Sunday (Sunday afternoon):
Fig 5 -- NAM 33-hour forecast sounding for OUN valid 21Z, Sunday, April 3, 2011.  From the HOOT website.
The red line is the forecast temperature of the atmosphere as we go up in height, and temperatures decrease to the left in the image above.  See that strong kink in the temperature profile that I circled?  Generally the profile is slanting to the left--the temperatures are cooling with height.  But in that one layer the temperatures abruptly warm with height.  This is a capping inversion--it prevents air near the surface from easily rising above the cap.  If it could, it would have very good convective potential--the CAPE value listed on the above sounding is 1716 J/kg.  Generally anything above 500 J/kg is good for strong thunderstorms.  And the higher the number, the more energy available for convection.

So how would we get storms when there is a cap?  We'd have to break the cap somehow.  There are three main ways of doing this:

  1. Cool the air above the cap
  2. Warm the air below the cap
  3. Have a cold front or some other strong lifting mechanism move in that can shove the surface air right through the cap regardless of its relative buoyancy.
Usually some combination of these does take place, though the second one becomes most dominant in the absence of a front moving through.  As the day wears on, the sun heats up the earth's surface more and more, causing the temperature at the surface to get warmer and warmer.  Also, turbulent mixing can bring warm air down from above the cap and cool air up from below the cap and sort of "even out" the temperature, making the temperature jump less abrupt.  This combination of things works to erode the cap--which is exactly what the NAM model forecasts to happen just three hours later at 00Z:
Fig 5 -- NAM 36-hour forecast sounding for OUN valid 00Z, Monday (Sunday evening), April 4, 2011.  From the HOOT website.
Just three hours later, the NAM no longer forecasts there to be a cap present!  The air near the surface is now warmer than all the air above it and parcels can rise much more easily now.  So according to this forecast, there definitely is potential for the cap to break down and for us to see some thunderstorms further south.  The OU/OWL WRF model shows a similar profile for 00Z Monday, lending further credence to this possibility:
Fig 5 -- OU/OWL WRF 48-hour forecast sounding for OUN valid 00Z, Monday (Sunday evening), April 4, 2011.  From the HOOT website.
The NAM soundings on the HOOT site stop at 00Z Monday, but the WRF model goes slightly further, and I want to use its soundings to show something.  All of these maps and soundings I've shown are at 00Z Monday, which is around 7 PM central time on Sunday night--right before sunset.  What happens if we go three hours later, to 3Z Monday (10 PM Sunday night)?  Here's the OU/OWL WRF forecast sounding for Norman at that time:
Fig 6 -- OU/OWL WRF 51-hour forecast sounding for OUN valid 03Z, Monday (Sunday evening), April 4, 2011.  From the HOOT website.
The cap has reappeared!  Why is this?  As soon as the sun goes down, the earth's surface really starts to cool off.  As it cools off, this in turn cools the air right above the surface.  However, the air in the low-levels of the atmosphere still retains its warmth from the day.  So, as the surface cools, so does the air right above it.  But above that, we still have very warm air.  Thus, another capping inversion sets up, where the air warms abruptly with height.  This prevents new storms from forming that draw on surface air for their energy, and also causes already-existing storms to elevate above the capping inversion.  When a storm elevates, it no longer draws in air from the near-surface layer of the atmosphere and instead must draw in air from above this layer.  This is in contrast to surface-based storms that draw in their warm, moist air from the surface layer of the atmosphere.  Since the capping inversion stops near-surface air from easily rising, any surface based storms that run into a capping inversion tend to lift above the inversion and start drawing in air from above that level.  They can still be severe, but usually elevated storms are not tornadic--a tornado has to extend to the ground and it can't do that (easily) if the storm is elevated.

Notice one other thing in the last two profiles I've shown--the low-level winds pick up in speed rather extraordinary between 7 PM and 10 PM (see the wind barbs between the surface and 700mb in the two soundings above).  This has to do with the nocturnal low-level wind maximum (sometimes associated with what is also called the low-level jet).  The details of why this happens are enough material for another whole blog post, so I'll just say that as the sun goes down (and as that capping inversion becomes re-established), the low-level winds tend to speed up rather suddenly.  This is good for severe storms as they need wind shear to maintaining rotating updrafts and produce hail, strong winds and sometimes tornadoes. However, this night time increase in winds occurs at the same time that the capping inversion shows up again, causing the storms to elevate.  This is why the time right around sunset is often critical if you are looking for tornadoes--you want the atmosphere to stay uncapped so you still have surface-based storms, but you also want the wind shear to increase.  Right around sunset sometimes the capping inversion hasn't quite started to creep in yet and sometimes the wind shear will really start kicking up and you'll see an increased risk of tornadoes around then.

However, with the increase in low-level winds comes an increased tendency for storms to form bow-echoes and have strong, straight-line winds associated with them, regardless of if they're elevated or surface-based.  We often see an afternoon full of isolated supercells and small storms quickly congeal together into big lines of storms after the sun goes down and the low-level winds kick up.


So, in summary of all this--models are coming into agreement that severe weather is possible across much of the central US.  We're virtually guaranteed to see storms further north from Iowa and Missouri into Illinois because the cold front should be moving through.  The cold front is such a strong enough forcing mechanism that regardless of if there is a cap, it can force enough lift to cause some strong thunderstorms.  Further south, the threat becomes more conditional in the afternoon as the models are forecasting a strong capping inversion at many places.  As we saw in the examples above, models are also showing that the cap may erode enough in the late afternoon and early evening that we could see some good storm development.  However, the cap is forecast to quickly become re-established as night falls, limiting the surface-based development and consequently the strong tornadic potential.  However, as winds pick up around sunset, the potential for lines of storms with strong winds increases.

We'll see how it all plays out tomorrow...

No comments:

Post a Comment