Monday, April 25, 2011

Squall Lines and Rotation on Radar

I've been away for a week--and what a week it has been.  Repeated rounds of tornadoes, including one that destroyed dozens of homes and part of Lambert International Airport in Saint Louis.

Unfortunately, it's not over yet.  The SPC convective outlooks currently have moderate risks of severe weather posted for the next three days.  It's round after round of severe weather.

I want to quickly look at the model forecasts to show why we aren't getting out of the woods yet.  Here is this morning's European model 500mb analysis.  I've drawn dashed lines through two troughs--one in the central US which is causing tonight's severe weather in the lower Mississippi Valley and the other just now impacting the Pacific northwest with a strong jet streak right behind it.
Fig 1 -- ECMWF 500mb analysis of winds (colors) and heights (contours) at 12Z, April 25, 2011.  From the HOOT website.
The forecast for 24 hours later, tomorrow morning, shows the central US trough moving eastward and becoming more negatively tilted over the lower Ohio River valley.  In the meantime, the Pacific northwest trough is beginning to cross the Rocky Mountains.
Fig 2 -- ECMWF 24 hour forecast of 500mb winds (colors) and heights (contours) at 12Z, April 26, 2011. From the HOOT website.
The strong jet streak behind the trough over the Rocky Mountains hints that that trough will continue to deepen as it exits the mountains and enters the plains.  The eastern trough seems to become absorbed back into the flow while the western trough really digs in by Wednesday.
Fig 3 -- ECMWF 48 hour forecast of 500mb winds (colors) and heights (contours) at 12Z, April 27, 2011. From the HOOT website.
There's even yet another shortwave trough approaching off the Pacific northwest again by Wednesday.  As we keeping having troughs move through every other day, the chances for severe weather will keep coming and coming.

But today I wanted to focus more on observing severe weather on radar, mostly in describing squall lines and bow echoes.  Tornadoes are one major element of severe weather.  However, the strong winds in squall lines cause tremendous amounts of damage too--and over a much larger area.  What causes the very strong winds at the surface in these lines of storms?

Let's re-visit the basic model of a thunderstorm. As a storm (or several storms) move from west to east, they are fed by warm moist air rising in the updraft region.  As that air rises, it cools and the water vapor condenses into rain.  The now colder, rain laden air sinks rapidly back down to the surface in the downdraft region.  Often the lift needed to get that warm, moist air initially rising is provided by convergence along a boundary like a cold front which helps move the storms along.  Think of the cold front like a big snow plow, moving along and forcing warm moist air up and into the storm as it plows forward.  Notice that the cold air sinking in the downdraft region sinks behind the cold front--this influx of cold air at the surface actually helps reinforce the cold front boundary as the cold downdraft air spreads out.  In this way, the storms can actually become self-sustaining as their own downdrafts spread out and lift up surrounding warmer, moist air, further feeding the updraft.
Fig 4 -- Schematic of a cross-section through an idealized thunderstorm.  The storm is moving to the right.
So we can see one reason why there are strong winds at the surface as the storm moves through--all that cold, rain-laden air crashing back down to the surface would make for gusty conditions indeed.  But what makes squall lines in particular so destructive?

Remember in my previous discussions about tornado formation how the low-level winds just above the surface tend to really increase in strength right around sunset.  This phenomenon is called the "nocturnal low-level wind maximum" or the "low-level jet" (these actually are different things, but the nocturnal low-level wind maximum is a kind of low-level jet...).  So let's say that we have a strong jet of winds aloft just off the surface, like what we often see around sunset.  What would that do to the storms?

Let's imagine a low-level jet of strong winds approaching the storm from the rear.  What happens as the storm interacts with the jet?  The rapidly sinking air in the downdraft region of the storm forces that jet of winds downward until it runs into the surface.  This is what tends to give squall lines their particularly powerful straight-line winds--a low-level jet of air that is pushed down to the surface.  In this case, the jet is often called the "rear inflow jet".  Not only does this interaction bring those very strong winds down to the surface, but all that air rushing in behind the storm can force the cold front or outflow boundary (that snow plow that is helping to lift up the warm-moist air in the updraft) to surge forward rather quickly.  Thus, the storms in the area where the low-level jet is intersecting the boundary will tend to rush forward faster than storms further from the center of the jet.  This is what creates the "bow echo" structure so often seen on radar--the storms "bow out" or rush forward where the strong winds of the center of the low-level jet are being forced to the surface, causing those particular storms to surge ahead.  No wonder that the strongest winds are often found on the leading edge of a bow-echo.
Fig 5 -- Schematic of a cross-section through an idealized thunderstorm interacting with a low-level jet flowing in from the left.  The downdraft forces the jet down to the surface, causing very strong winds at the surface.  Furthermore, this strong air at the surface causes the storms in the area right in the middle of the interaction with the jet to surge forward, forming a bow echo.
So that's the basic reasoning behind bow echoes and why they have such strong winds.  It's all because of the downdraft bringing down a jet of stronger winds from just above the surface.  There are lots of nuances and details to how this structure works, but that's the general idea.  For more detailed descriptions of these structures (in more technical terms) I would suggest papers like Smull and Houze, 1987  , Houze et. al 1989  or Biggerstaff and Houze 1991.So how does this look on radar?

Here's a reflectivity image of a squall line/bow echo from last week as it approached Champaign/Urbana, IL:
Fig 6 -- KILX 0.5 degree base reflectivity image from 0002Z, 4/20/2011.
Can you imagine the structures I described above?  Rain falls in the downdraft region, so we know where all of the downdraft area is--all the high reflectivity areas of red and yellow behind the leading edge of the line.  Notice the bowing structure of the leading edge of the reflectivity.  You can imagine that there are strong winds being forced down in that rain-filled downdraft area, causing the leading edge to surge forward.  We can see this on the base velocity image, too.
Fig 7 -- KILX 0.5 degree base velocity image from 0002Z, 4/20/2011.
Remember red colors indicate air moving away from the radar while green colors indicate air moving toward the radar.  In this image, the radar is toward the top of the image to the left of the center.  I've kept the schematic arrows reminding us of how the low-level jet air is being forced down to the surface.  Notice that there is a maximum of air moving away from the radar (bright red colors) right behind the leading edge of the line. This is where those upper-level winds are really being dragged down to the surface.  Immediately in front of the line is air moving toward the radar--in the updraft region.  The strong convergence between the air being dragged down to the surface and the warm air out in front is helping to force that warm moist air up ahead of the storms and keeping them going.

I wanted to zoom into the velocity image for a moment to talk more about properly interpreting radar velocity images.  Remember--the radar can only measure air moving toward or away from it--so if we try to determine wind direction, those are our only two options.  I talked in a previous post about how strong inbound air (green colors) right next to strong outbound air (red colors) can be evidence of rotation.  But here we have strong reds and greens right next to each other along the leading edge of the line.  This is NOT rotation--this is convergence.  Here's a closer view:
Fig 8 -- Zoomed in version of figure 7.  The circled areas indicate potential rotation.
You always have to keep in mind where the radar is located to determine what direction the winds are going with respect to the radar.  In the image above, the radar is off the upper-left corner.  So, when I start drawing arrows to show wind direction, all of my arrows in the green section must point directly toward the radar.  Likewise, all my arrows in the red section must point directly away from the radar.  When we do this, we see that the winds in the southern part of the line approach each other head-on--this is convergence, not rotation. For there to be rotation, we would need the arrows to be moving past each other showing spin--not right at each other showing convergence.  However, there are a few small areas where we can indeed see some rotation.  Notice the areas further north on the line where I put blue circles.  There are small "chinks" in the otherwise straight line--small areas where if you were to draw those arrows, they would point past each other instead of directly head on.  Can you see how this would happen?  It's kind of difficult to see at first.  But, with some practice, these little "chinks" in the line where the velocity arrows would be pointing past each other instead of right at each other stand out and can provide evidence of embedded rotation.

Here's another example.  This is from this afternoon's squall line moving through Arkansas.  Here's the reflectivity image.
Fig 9 -- 0.5 degree base reflectivity from KLZK, 0101Z April 26, 2011
Once again we see a relatively nice bowing line on reflectivity.  Can you imagine a jet of air descending down right in the middle of that bow and pushing it forward?  I hope so by now.  Notice that I circled a little area where high reflectivity returns seem to stick out a little bit ahead of the line. What would cause a very small area of the downdraft to shoot out even further ahead of the front of the line?  Could there be rotation there, where rotating air is wrapping the downdraft around and pushing it out ahead even faster than the descending low-level jet air?  If so, then this little reflectivity bump could be a small hook echo--we should investigate.

Here's the base velocity image corresponding to what we see above.
Fig 10 -- 0.5 degree base velocity from KLZK, 0101Z April 26, 2011
Remember--all the arrows I draw have to be pointing directly toward or directly away from the radar.  You can see the location of the radar as the small blue dot labeled "KLZK" just above and to the left of the center of the image.  Because we're so close to the radar, you can see the the inferred wind directions change depending on where they are in relation to the radar.  Nevertheless, we see that almost everywhere there is just implied convergence--evidence of strong winds and strong storms.  However, I left the circle on our area of interest.  Notice that there is indeed a little "chink" in the line separating the inbound and outbound (the red and green) velocities there!  Let's zoom in even closer:
Fig 11  -- Zoomed in figure 10,  Slight rotation is implied in the circled area.
You can see the location of the radar in the very upper left corner of the image.  If we draw our wind velocity arrows with that in mind, we see that right where that "chink" is in the circled area, the arrows can go past each other instead of hitting each other head-on.  This does imply some weak rotation there instead of just straight convergence.  So...probably based on that reflectivity bump and the fact that the velocity image does show some potential for rotation, the NWS put that area under a tornado warning.

You can see that not all bright reds next to bright greens on radar velocity images mean rotation--often they mean convergence or divergence.  You must keep in mind where the radar is located and what that means for the implied wind directions to accurately assess the radar velocity image.  The simple rule is this--draw a line directly out from the radar location to your area of interest.  If the line separating the reds and greens is perpendicular to the line drawn straight out from the radar, then it is implied convergence or divergence.  If the line separating the reds and greens is parallel to the line drawn straight out from the radar, it could imply rotation (particularly if it's a very small, localized region).  Here's the above image with that guideline applied to it:
Fig 12 -- Figure 11 annotated to describe areas of convergence versus rotation.
See the difference between the orientation of the red-green boundaries?

Here's another, more advanced but still cool thing that we can see on radar.  With certain radar viewing software, we can examine vertical cross-sections of the reflectivity and velocity fields shown by the radar and compare them to the idealized schematics I drew above.

Here's a reflectivity cross-section through the heart of the squall line in figures 9-12 as it approached Memphis this evening:
Fig 13 -- Base reflectivity cross-section from KNQA at 0313Z, April 26, 2011.
Remember, this is a cross-section--I basically just took a vertical slice right through the bow echo to show its vertical structure.  So, up on this image is actually up--height increases in the vertical with the ground at the bottom.  The radar is located just off the right side of this image.  I drew in the arrows showing the rough locations and orientations of the updraft and downdraft.  It's difficult to really see these structures on reflectivity, but we know that the heaviest rain is in the downdraft region.  So, the big column of dark reds and even pinks near the surface is associated with the downdraft.

Things get more interesting when we look at the velocity cross-section:
Fig 14 -- Storm-relative velocity cross-section from KNQA at 0313Z, April 26, 2011.
Now we can really see a lot of the structures I was describing.  Remember, the radar is located just off the right side of this image.  So--red colors show air moving away from the radar and green colors show air moving toward the radar.  You can see that well-defined column of red colors showing air moving up and away from the radar on the right side of the image.  This shows the updraft region very well--air moving up and back into the storm.  Furthermore, we can trace the low-level jet, which is marked by air heading toward the radar.  We can see a ribbon of green colors, indicating air moving toward the radar, that runs from the left the side of the image before coming down to the surface right behind the updraft.  This very nicely shows the low-level jet being pushed down to the surface by the downdraft.  You can see the strong convergence along the leading edge of that jet forced to the surface as it runs into the updraft (I marked this with the little cold front symbol at the bottom of the image).  That convergence is helping to lift the warm air up into the updraft and keeping the storm going.  Pretty amazing that you can see all these structures on radar.

Hopefully you can now look at squall lines and bow echoes on radar and understand a little bit more of what's going on.  The powerful, straight-line winds associated with bow echoes have very distinct radar signatures that can provide advance warning of their approach.  Furthermore, sometimes rotation can be embedded in that leading line, but it takes a keen eye to spot the velocity signature corresponding to that.


  1. I've been into meteorology as a sort of hobby since I was 5. Still enjoy dabbling in it and recently found a way to access the better radar images in almost real time. So I've been doing some digging into things like base velocity.

    Very informative article. Learned quite a bit on how to read those suckers. Thanks again for all the info!

  2. Thanks a lot ive been a weather finatuc since I was 4 and have never understood those red and green colors in the map when the meteorologists read them, now I have a crystal clear understanding even at age 16 now.