Diffluence..?

[Chris C Sanner]

Here's a question of a concept I have a hard time grasping for some reason. So let's just start at the bottom as I'm sure there are some more with questions on it as well. What exactly is it? How exactly would it help or hurt a chase day? What do you look for in the models to tell where/how much, etc.?

 

[Chad Cowan]

Diffluence (or diffluent flow) in the mid and upper troposphere triggers the air below to rise to fill the space of the diverging air. This rising motion is very favorable for convective development.

Take a look at the 500mb chart below. The wind barbs over the panhandles are parallel, while the barbs over Goodland and Wichita are nearly perpendicular. The air begins to 'fan out' or diverge over W KS creating a void that is filled by the air drawn up from below.

 

[Bill Tabor]

Difluence is a component of divergence. Keep in mind they are not the same. Divergence is composed of directional difluence and speed divergence. While difluence is typically seen as the horizontal directional spreading of air in the atmosphere (as seen in chart above) the lift component and benefit for severe storm development may be cancelled if speed convergence is taking place at the same time. If in a jet stream at a particular level ~ say 500mb contains difluent motion but speed is remaining constant or moving faster then mass is being removed from the atmosphere creating low pressure and therefore causing lift (vertical motion) at lower levels. In those cases it is one item to look for which is favorable for severe weather.

When examining upper level jet flows, for instance 500mb or above, the areas most favorable for severe weather development when viewed as a stream of air (streamwise regarding flow direction) are the right entrance and left exit regions where divergence and vertical motion is maximized.

 

[Jeff Snyder]

As Bill noted, it is very important to realize that difluence does NOT necessarily imply divergence. Indeed, divergence is a function of difluence (and "directional" component, of sorts) and a speed component. Oftentimes, you'll see difluence downstream of the jet streak, right where you also see speed convergence (and vice versa for the area immediately upstream of a jet streak). Of course, the situation is much more complicated than this thanks to various ageostrophic accelerations. For example, ageostrophic divergence/convergence in the four quads of a "textbook" straight jet streak results in a transverse circulation that we've come to know (div in the right-entrance and left-exit regions; conv in the left-entrance and right-exit region).

Let's put this in context of a classic severe weather outbreak in the Plains. Such setups are often characterized by an upper-level trough over the Rocky Mountains and a jet streak "nosing" out onto the Plains states. The area downstream of a trough and upsteam of a ridge (or, the area in the Plains, if there is a trough to the west and a right to the east) will be an area where there is ageostrophic divergence associated with the change in the sign of curvature between the trough and ridge. In addition, there will be ageostrophic divergence in the left-exit region (the polar side) of the jet streak. Depending upon many other elements, the upper-level div in the left exit region (likely resulting in large-scale upward motion) and upper-level convergence in the right exit region (likely resulting in large-scale subsidence) will enhance the low-level jet over the Plains (the sinking motion on the south side may yield surface divergence to the south, while upward motion on the north side may yield surface convergence, thus the enhancement of the LLJ). So, there could be significant large-scale ascent over portions of the plains associated with both ageo curvature divergence and ageo div associated with the jet streak. Add in a vort max that is moving onto the Plains (where you may see differential cyclonic vorticity advection), and you could see significant mid level ascent (upward motion = cooling). Again, I have to say "could be" and "may yield" since this is only an idealm simple analysis.

It's also important to remember that you should use 200-300mb charts if you want to equate divergence = upward motion in the low levels and 1000-850mb charts if you want to equate convergence = upward motion above. If you try to discern low-level upward motion with a 500mb chart, you may have problems. For example, what happens if the "air" that's being evacuate in an area where you think there will be divergence at 500mb is not resulting in low-level upward motion but instead by upper-level downward motion? Static stability has a significant impact on the amount of vertical motion experienced as a result of differential vorticity advection (DPVA/DNVA) and thermal advection (CAA/WAA) which further complicates this.

If you'd like to learn more jet streak dynamics (very important for severe weather forecasting), I strongly suggest you watch the COMET MetEd module on jet streak at https://www.meted.ucar.edu/loginForm.php?urlPath=norlat/jetstreaks# . You'll have to register first, but there are many other very informative MetEd modules that you may want to view as well.

 

[Tim Vasquez]

Jeff's elaboration on this is well worth considering... difluence back in the 1960s and 1970s was widely cited as a pattern seen with most tornado outbreaks; a classic item from your Colonel Miller checklist. However the difluence is usually negated by speed convergence (look upstream and you'll often see 50-100 kt winds approaching). It's good to know what it is, but in practice it's safer to turn to a computer-solved divergence field, like this... or better yet use some of the specialized diagnostics mentioned above.

Tim

 

[Bill Tabor]

Yes, let me add that Jeff is correct in his preferential use of 200-300mb for jet analysis as opposed to 500mb. I sometimes use 500mb, but as he mentions it isn't always a safe bet if the area above does not also have upward motion.

 

[Rich Thompson]

Yes, let me add that Jeff is correct in his preferential use of 200-300mb for jet analysis as opposed to 500mb. I sometimes use 500mb, but as he mentions it isn't always a safe bet if the area above does not also have upward motion.

What Bill's getting at is this - upward motion is related to *differential* divergence (increasing with height, same as convergence decreasing with height). Divergence at any one level doesn't mean much, and difluence is an unreliable indicator of divergence.

As Jeff mentioned, static stability (lapse rates) play a big role in the magnitude and distribution of vertical motion. Steep lapse rates promote stronger vertical motion on smaller scales, while stable lapse rates promote weaker ascent on broader scales. Then you have to consider the vertical distribution of static stability when trying to link mid-upper ascent with the boundary layer. It all gets pretty complicated in a hurry, and there are no easy fixes.

I agree with Tim - just look at calculated divergence, or don't bother. The SPC web site's upper air charts have divergence calculated at 300 mb, FWIW. On the large scale, divergence is related to vorticity advection, so it's almost easier to infer divergence by looking for regions of vorticity advection. Vorticity advection increasing with height is related to height falls and ascent in QG land.

Rich T.

 

[Glen Romine]

On the large scale, divergence is related to vorticity advection, so it's almost easier to infer divergence by looking for regions of vorticity advection. Vorticity advection increasing with height is related to height falls and ascent in QG land.

I would note that while a vorticity maximum typically underly a jet streak (a short wave, or vorticity maximum, is a localized region of enhanced temperature gradient - which through thermal wind relation must have stronger winds above the stronger temp gradient), in which case there is similar information content. But, upper level divergence associated with large scale flow curvature is not necessarily associated with differential vorticity advection, unless there happens to be a shortwave passing through the larger scale pattern. So in short, you may still have upper level divergence in the absence of a shortwave.