Friday, April 8, 2011

Diagnosing a Day-Three Moderate Severe Risk

I shouldn't necessarily say "upcoming"--there are some ongoing severe storms in northern Oklahoma and eastern Kansas even late this Friday evening:
Fig 1 -- KICT 0.5 degree base reflectivity from 0345Z, April 9, 2011.
Several reports of large hail were associated with these storms as they moved through the area.  No tornadoes reported with these storms.  There were possible tornado reports with another group of storms in the Kentucky/West Virginia area on Friday afternoon, however.

But let's get right down to it.  The SPC has another slight risk of severe weather tomorrow for two regions--one in the upper midwest and the other in the Ohio Valley down through the Carolinas.  The big buzz as of late, though, is for the day 3 moderate risk issued on Friday for Sunday.
Fig 2 -- SPC day 3 convective outlook issued 07346Z, Friday, April 8, 2011.
As Patrick Marsh noted on his blog, day three moderate risks are rather rare--in the past ten years there have been only eight day-three moderate risks issued by the SPC.  To issue such a significant risk from three days out means that the SPC has high confidence that there will be severe weather in that particular region during that time period.  Let's see what the latest models have been saying.

To get severe weather (in general) there are five basic ingredients (at least, this is how I was always taught to think about this:

  1. Shear
  2. Instability
  3. Lift
  4. Moisture
  5. "Exhaust"
Why do I put them in that order?  Because when you make an acronym out of them, you get "SMILE".  You can tell I was taught by veteran storm chasers...

So let's start looking at these particular parameters, though not in that order.  The "Exhaust" parameter is supposed to refer to upper-air features that can evacuate air out at the top of a thunderstorm from fast-rising convective plumes.  I prefer to think of this ingredient as "upper-level support" for a more wide-scale convective situation--i.e., I'm looking for divergence aloft.  Here's the current synoptic pattern at 300 mb (the upper troposphere):
Fig 3 -- 00Z analysis of 300 mb winds (colors) and geopotential height (contours) for April 9, 2011.
There's a deep, positively-tilted trough over Nevada and California on Friday evening.  The eastern US has a generally zonal pattern with a weak ridge across the central US.

So what's going to happen aloft?  The models are in agreement that the deep trough will be slowly moving across the country over the next two days.  What's remarkable about this trough right now is that it's staying very deep but is still advancing eastward fairly steadily.  Often when we get really deep troughs they have a tendency to "cut-off" from the main flow and just spin around over one area for a while without really moving. But this one seems to be on the move.  Here's the NAM model 300 mb wind forecast for Saturday evening:
Fig 4 -- NAM 24 hour forecast of 300 mb winds (colors) and geopotential height (contours) for 00Z, Sunday (Saturday night), April 10, 2011.
The 300 mb trough has moved over the Rocky Mountains and a strong jet streak is forecast to build on its eastern side.  I'm not too impressed with the flow to the east of that, though.  There is a weak(er) east-west jet streak across the northern Great Lakes, but otherwise there's generally weak flow coming down across the Ohio Valley and the Carolinas.  This makes me think that the severe threat tomorrow for both those areas is going to be somewhat conditional--there doesn't look to be widespread upper-air support in the models.  However, that strong jet streak on the leading edge of the trough is definitely going to start driving things down below.  Here's the NAM surface forecast for Saturday evening:
Fig 5 -- NAM 24 hour forecast of surface temperature (barbs), mean sea-level pressure (contours) and winds (barbs) for 00Z, Sunday (Saturday night), April 10, 2011.
As we would expect, under the divergent exit region of this jet streak (which has a slight cyclonic curvature to it) we see lower pressure developing.  Pressure is going to be falling all day Saturday in the central plains as that jet streak approaches from the west.  With falling pressure, winds are going to pick up trying to fill the ever-increasing pressure difference.  As such, you can see that strong southerly winds are being forecast across a broad swath of the south-central part of the country.  Warm (and moist, as we'll see in a minute...) air is being brought up north with these winds.  The relatively weak upper-level pattern across the eastern US today and tomorrow has left a relatively "unforced" flow at the surface, which is easily distorted by the lowering pressure in the central plains.  This is allowing for that impressively large warm sector to develop.  We can see some boundaries forming--there does appear to be a cold front shaping up in this forecast across northern and western Nebraska.  The shift of winds from southwesterly in west Texas to more southerly across Oklahoma and eastern Texas hints that here may be a dryline forming there.  We can check the dewpoint forecast for that:
Fig 6 -- NAM 24 hour forecast of surface dewpoint temperature (colors) and winds (barbs) for 00Z, Sunday (Saturday night), April 10, 2011.
There is indeed a dryline forecast to be present at that location.  It isn't the sharpest dryline I've seen, but there's a clear division between moister and drier air forecast to set up from central Kansas down through western Oklahoma and into Texas.  The amount of moisture being brought northward is very rich to the east of the dryline.  Note the wind barbs--the flow across the entire central part of the country is forecast to be coming right out of the Gulf of Mexico.  This is a lot of moisture--with dewpoints in the 70s being forecast in northern Missouri, southern Iowa and southern Illinois.

The structure of this low-pressure center is also kind of unique in these plots.  If you go based on the structure of the moisture field, the main low-pressure center seems to be somewhere in northeastern Nebraska at this time.  Futhermore, it looks like the dryline extends all the way to the low-pressure center.  The cold front that we saw on the surface temperature forecast can also be seen in the moisture field as the pronounced wind shift between northerly to southerly winds stretching through western Nebraska and into northern Colorado.  But look at the location of the cold front on the dewpoint forecast map--the cold front is advancing into the dry air behind the dryline.  This means that, though the cold front would provide lift, there isn't much moisture in the air it would be lifting.  As such, the cold front isn't going to be much of a player on Saturday.

That upper-level trough is forecast to move slowly--by Sunday afternoon the NAM forecast has it positively tilted out over the great plains:
Fig 7 -- NAM 45 hour forecast of 300mb winds (colors) and geopotential height (contours) for 2100Z, Sunday, April 10, 2011.
This is still forecast to be a very deep trough--it extends all the way from Canada down to Mexico.  Furthermore, the jet streak on its leading edge is very well-organized in this forecast, with 140 knot winds in a rather compact area.  With such a strong, straight jet, we'd have to assume that the cold front has begun to assert itself and has moved down into the central plains.  Also, based on their being divergence in the left exit region of a straight jet streak, we'd expect to see the surface low somewhere over northwestern Iowa or southern Minnesota.  What does the NAM say?
Fig 8 --  NAM 45-hour forecast of surface temperature (colors), mean sea-level pressure (contours) and winds (barbs) for 21Z, Sunday, April 10, 2011.
Well--the low-pressure center is not as organized as I would have thought.  The NAM seems to have some difficulty placing the low-pressure center, but based on the swirl in the wind field our initial guess of northwestern Iowa does seem reasonable.  The cold front is right about where we expected it to be with that jet streak aloft, however.  But by Sunday afternoon, it's clear that the warm sector has reached at least all the way into southern Wisconsin, with a cold front roughly from Omaha down through central Kansas.  It's difficult to figure where the dryline would be and where the cold front really is once we get further south from there.  The dewpoint forecast should help with that:
Fig 9 --  NAM 45-hour forecast of dewpoint temperature (colors) and winds (barbs) for 21Z, Sunday, April 10, 2011
The moisture field shows a dryline actually located a bit further east than I would have guessed--located from the Kansas City area down through central Oklahoma and into central Texas.  Once again, the cold front still seems to be advancing into the dry air behind the dryline--so the cold front still looks to be playing a minor role in forcing lift for convection.  Yet there is a moderate risk out for eastern Iowa, northern Missouri, northwestern Illinois, southwestern Wisconsin and southeastern Minnesota for this time.  Where's the lift?

Let's re-evaluate our conditions for severe weather in that area with what we know so far (out of order):

  1.  Moisture--This one seems to be very well-established.  Dewpoints at the surface are forecast to be in the 60s or greater all the way up to southern Wisconsin in the forecasts.  That's definitely the kind of moisture we would be looking for to get severe weather.  Moisture is not an issue.
  2. Exhaust/Upper-level support--The 300mb winds for Sunday afternoon do have the broad left-exit region of a jet streak over the moderate risk area on Sunday afternoon.  That should provide enough divergence aloft to support thunderstorm growth below.
  3. Wind shear--Haven't really touched on this one much yet, but we can check it quickly.  Note that the surface winds in the warm, moist sector at this point are forecast to be rather strong--20 knots out of the south-southwest in the moderate risk area. The winds are somewhat lighter further north, but I don't really trust the NAM's winds in that "area of ambiguity" surrounding the low-pressure center.  Let's compare these winds with the forecast 850mb winds (a little ways off the ground):
Fig 10 --  NAM 45-hour forecast of 850 mb winds (colors) and geopotential height (contours) for 21Z, Sunday, April 10, 2011.
Note that the winds are more southwesterly just a bit off the surface over the moderate risk region.  This points to directional wind shear, at least.  Winds are also somewhat faster than at the surface--30-40 knots across the moderate risk region.  So there's good speed shear as well.  However, remember in one of my recent blog posts I talked about how the low-level winds (but not the surface winds) tend to pick up right after sunset.  Here's the forecast 850 mb winds for 6 hours later at 03Z (around 10 PM CDT).
Fig 11 --  NAM 51-hour forecast of 850 mb winds (colors) and geopotential height (contours) for 03Z, Monday, April 10, 2011
Look at how much the wind speeds increased as soon as night fell!  Now we're looking at west-southwesterly winds at 50-60 knots in the low levels of the atmosphere right above the surface in southern Wisconsin.  That's a whole lot of low-level wind shear.  So, I definitely think that wind shear will be sufficient to produce severe storms in the moderate risk region.


But what about the last two--lift and instability?  Some would argue that these are the most important ingredients to look for.  Instability seems to be a given.  Notice how warm the air was at the surface in the forecasts above.  Some people in southern Wisconsin and northern Illinois are forecasting highs in the mid 80s on Sunday.  Combine this with dewpoints in the 60s and it becomes very hard to consider the situation "stable".  Let's check the forecast NAM soundings from Earl's Skew-T page to see what we're looking at.  Here's the forecast sounding for Davenport, IA (roughly in the middle of the moderate risk region) for 21Z Sunday (Sunday afternoon):
Fig 12 -- NAM 45 hour forecast souding for KDVN, valid 21Z, Sunday, April 10, 2011.
I apologize for the small type and the cluttered diagram--if you click on it you can see the full-sized version and actually read things.  Note the CAPE value given of 2702 J/kg.  A very large amount of convective available potential energy.  But can this instability be tapped into?  Notice that there is still a capping inversion present at around 800-850 mb.  Remember that this capping inversion is a layer where the temperature of the environment actually warms with height instead of cooling like normal.  We can see that abrupt bump in the temperature curve right at that 800-850 mb layer.  So we do have a capping inversion to deal with that will keep a lid on tapping into that instability.  However, let's look at the forecast sounding for three hours later at 00Z (around 7 PM CDT):
Fig 13 -- NAM 48 hour forecast souding for KDVN, valid 00Z, Monday (Sunday evening), April 10, 2011.
The cap is no longer present!  So this forecast sounding has an uncapped profile with 2871 J/kg of CAPE.  Very unstable.  Very dangerous.  Particularly with all that wind shear...

Just for comparison, let's look at a point further south like Saint Louis:
Fig 14 -- NAM 48 hour forecast souding for KSTL, valid 00Z, Monday (Sunday evening), April 10, 2011.
Not as impressive.  Note that the temperatures don't cool as rapidly with height as they did up in Davenport.  There seems to be warmer air in the mid- to low-levels of the atmosphere that is not present further north.  There's still CAPE listed--at some 1532 J/kg it's pretty significant, too.  But the level of free convection (the level you'd have to lift surface up to before it could tap into that instability) is up at around 625 mb.  That's an awful long distance to have to lift air before it freely convects.  Particularly without a cold front in play to do the lifting for you.  The difference in the atmosphere further south partially explains why the moderate risk area was placed further north.

So what about that lift?  I believe that's the most conditional part of this setup, actually.  We've seen that the model says there will be upper-level support, wind shear, moisture, and instability.  All it takes is a little convergence at the surface to provide the lift necessary to start thunderstorm development.  This is particularly true with a very unstable and uncapped profile like we see in the forecast for Davenport.  Since some thunderstorms are expected in the region on Saturday, residual outflow boundaries and other perturbations in the wind field from the Saturday storms may linger on into Sunday and provide focal points for new convective development then.  With so many factors favoring thunderstorm development, it's virtually certain that storms will form.  In fact, with wide-spread vertical motion favored by the divergence aloft, we could see storms firing up all over the place.  It will be interesting to see how this event is organized.

Without an organized forcing mechanism like a cold front, the storms that form will have a greater tendency to remain surface-based, drawing warm moist air directly from the surface layer.  Combine this with the strong low-level wind shear and you have the ingredients for tornadic thunderstorms.  This also justifies the heighted "moderate" severe risk that the SPC has put out there.

More updates will come as this event draws closer.

No comments:

Post a Comment