04/27/14 Questions

[David Williams]

These questions have been repo status from the miscellaneous discussion for the same day.

So I was anticipating storms to fire off the dryline in eastern Oklahoma. My chase partner and I waited forever for the dryline to make its way east. Then, some things happened that I don't entirely understand, and I was hoping to get some clarification on:

1. The dryline was moving (I think) slower than anticipated across OK. I thought it was going to be the firing mechanism for storms that developed on far eastern OK and pushed into AR. Why was the dryline moving so slow? Now my guess is that it ran into so much deep moisture with such a stout cap that it was having difficulty making progress; however, I haven't been able to find soundings from yesterday to see if the moisture was reaching 5000 ft. (Interestingly enough, I learned about drylines being slowed by deep moisture from Tim Vs article "Dryline Magic" that I read on Saturday).

2. The first storm to fire off the dryline and track into OK was way south on the on the TX border near Sherman. This guy started looking all big and bad, but after a while I noticed that he had halted the dryline, decoupling himself from it and moving away NE. So, why did he stop the dryline? And, why did he and the small severe storm to the north die out once they left the dryline?

3. As the dryling tracked closer, I was watching towers go up and die all over the place. Was this due to the strength of the cap? Storms being sheared off? The slow momentum of the dryling? Or, another reason?

4. The only place where the dryline kept advancing was over northern OK. Hark! Two cells are produced here, both go severe quickly, and one produces the Quapaw and Baxter tornadoes. So, were these storm able to go severe quickly and produce the tornadoes because they were anchored on the dryline?

5. If the dryline did make it into AR, would we have seen some other crazy storms/tornadoes especially in the HIGH Risk area?

Thanks for any feedback!

Sent from my SAMSUNG-SGH-I337 using Tapatalk

 

[Jeff Duda]

1. The dryline was moving (I think) slower than anticipated across OK. I thought it was going to be the firing mechanism for storms that developed on far eastern OK and pushed into AR. Why was the dryline moving so slow? Now my guess is that it ran into so much deep moisture with such a stout cap that it was having difficulty making progress; however, I haven't been able to find soundings from yesterday to see if the moisture was reaching 5000 ft. (Interestingly enough, I learned about drylines being slowed by deep moisture from Tim Vs article "Dryline Magic" that I read on Saturday).

That sounds like a plausible hypothesis to me. I believe that excessively deep moisture may indeed hinder dryline propagation since they move by mixing dry air from above the PBL to the surface. If there's no dry air to mix, then the DL wouldn't be expected to move much.

By the way, you can check the SPC site's "observed sounding analysis" page (http://www.spc.noaa.gov/exper/soundings/) for soundings that are no more than a week or so old. After that, check the severe storm events page (http://www.spc.noaa.gov/exper/archive/events/).

2. The first storm to fire off the dryline and track into OK was way south on the on the TX border near Sherman. This guy started looking all big and bad, but after a while I noticed that he had halted the dryline, decoupling himself from it and moving away NE. So, why did he stop the dryline? And, why did he and the small severe storm to the north die out once they left the dryline?

While drylines tend to move in a macroscopic sense due to mixing, little bulges can form due to advection of particularly dry, cold, wet, or warm air masses near the boundary. Thunderstorms put out rain cooled outflow. This outflow is typically much cooler than the surrounding unperturbed air, but is also more moist. Thus, the outflow coming out the back of the storm probably nudged the dryline backward, but only in the vicinity of the cold pool. Storms move due to the average wind in the depth of the storm, which typically extends from not far above the ground ( a few hundred meters perhaps) to the tropopause (10-15 km depending on time of year and how much CAPE there is). The mid and upper level flow was decidedly out of the southwest, so the storms were pushed to the NE and off the dryline (since the dryline wasn't moving).

Determining whether a storm will live or die is a complex process. Just getting a storm to form is a bit easier to determine - one only needs a lifting/triggering mechanism in the presence of instability. However, once a storm moves away from the trigger, it may not become well established if the atmosphere away from the trigger is insufficient to keep it going. You need CAPE and not too much CIN to keep a storm going. The amount of CIN that becomes "too much" depends on how well-established the storm is. Storms that have been going for awhile possess a constant stream of updraft air. The width of the updraft and the amount of dry mid-level air that entrains into the updraft partially control whether further entrainment will kill off an existing storm. Even in the presence of significant CIN, however, a storm that has a strong, consistent, stable updraft can provide enough forcing and momentum to get air parcels to their LFCs and through the stable layer in which the CIN exists. Newer storms tend to have narrower and more heterogeneous updrafts, and those can be cut off by entrainment or by lower amounts of CIN.

Keep in mind the following: air mass must be conserved. Meteorologists use what is called the continuity equation to express this. It basically says that if you have 2D divergence (along 2 of the 3 spatial axes), then you will have divergence along the third one. When air goes into the updraft, other air parcels must move in to replace the air lifted into the updraft. This creates horizontal convergence below the updraft. By the continuity equation, you must have vertical divergence (which means vertically accelerating air since air can't move through the ground) going on.

3. As the dryline tracked closer, I was watching towers go up and die all over the place.

The cap probably didn't strengthen. However, if you were getting late into the day, surface temps may have been cooling and thus you had increasing CIN. It's also possible the flow near the dryline became less convergent, thus providing less lift. These are some of the things that kill storm chances late in the day. Another thing that can do this is if a wave passes by and the downward branch of the wave is over the forcing, thus resisting lift. That's usually easy to see by looking at charts, though. I don't recall if this happened this day, and I doubt it did.

4. The only place where the dryline kept advancing was over northern OK. Hark! Two cells are produced here, both go severe quickly, and one produces the Quapaw and Baxter tornadoes. So, were these storm able to go severe quickly and produce the tornadoes because they were anchored on the dryline?

I don't know this for sure, but I seem to recall seeing low-level shear was better over NE OK/E KS/W MO than it was over E-SE OK/W AR. CAPE was about the same. There also might have been some mesoscale enhancement to the surface flow near the DL to the north since that area had been worked over by a morning MCS, thus providing enhanced lift to get storms to fire there.

5. If the dryline did make it into AR, would we have seen some other crazy storms/tornadoes especially in the HIGH Risk area?

The traditional southern Great Plains dryline tends to produce lots of cellular storms (as opposed to squall lines). So in that sense, yes, had the dryline made it into AR, there's a chance more tornadoes would've occurred. However, I think the big failure mode (relatively speaking) on that day was the fact that the warm front across N AR had been halted by persistent convection throughout the morning and afternoon. Remember how I told you that storms put out cold and moist outflow? MCSs tend to be really good at reinforcing warm front positions by hindering their retreat through persistent storm outflow over a widespread area. I think the models forecast the warm front to lift much farther north by 00Z or so than it actually did. That big storm that produced the Mayflower-Vilonia tornado struggled mightily and pretty much died once it crossed the warm front northeast of Little Rock. It did continue to produce tornadoes as it interacted with the front, however. I bet enhanced horizontal vorticity associated with the warm frontal circulation played a big role in getting those later tornadoes to form. Had the storm not been so well established once it began to interact with the warm front, I doubt it would've produced any more tornadoes.

 

[David Williams]

Wow Jeff, awesome play-by-play explanation! I totally appreciate the time you took to respond. In fact, it was so good that it inspired two clarifying questions and another big question; but before I throw those out, I wanted to mention that I looked at the spc sounding analysis page you recommended and sure enough the temp and dew point lines on the 18z sounding for Joplin looked like a pipe from near surface all the way up to about 200 mb!

(I don't have a handle on quoting other people's posts yet, so I'll do this the old fashion way)

Jeff says: "I don't know this for sure, but I seem to recall seeing low-level shear was better over NE OK/E KS/W MO than it was over E-SE OK/W AR. CAPE was about the same. There also might have been some mesoscale enhancement to the surface flow near the DL to the north since that area had been worked over by a morning MCS, thus providing enhanced lift to get storms to fire there."

Clarifying Question: When you say, "mesoscale enhancement to the surface flow near the DL to the north since that area had been worked over by a morning MCS, thus providing enhanced lift to get storms to fire there" does that mean an outflow boundary?


Jeff says: "I think the models forecast the warm front to lift much farther north by 00Z or so than it actually did. That big storm that produced the Mayflower-Vilonia tornado struggled mightily and pretty much died once it crossed the warm front northeast of Little Rock. It did continue to produce tornadoes as it interacted with the front, however. I bet enhanced horizontal vorticity associated with the warm frontal circulation played a big role in getting those later tornadoes to form. Had the storm not been so well established once it began to interact with the warm front, I doubt it would've produced any more tornadoes."

Clarifying Question: So am I right to believe that as storms cross a warm front they actually have a small window where interacting with the warm front actually strengthens them, or enhances the chance of tornadogenesis? But, once they sufficiently pass the warm front, they die off rapidly?


Jeff says: "Determining whether a storm will live or die is a complex process. Just getting a storm to form is a bit easier to determine"

Big Question: What is the best way to determine if a storm will live or die. I have been wondering about this for a long time. Is it best to really study the atmosphere (with real time surface obs and models) the morning of the chase so that you know exactly what the atmosphere in front of the storms is supposed to be like and therefore you will know which storms to pick? Or is there a way to make educated guesses while you are actually chasing in order to know what storm to pick? Or, am I thinking too microscale on this, and once you have chosen a target area, it becomes a crap shoot and going tail-end-charlie is just the safest bet?

 

[Mike Johnston]

Remember, too, that there was also a cold front involved. By looking at some of SPC's mesoscale discussions, it sounds like the cold front overtook the dryline around mid-day in south central KS and north central OK at least. When a cold front overtakes the dryline, it can render the dryline impotent and irrelevant as a trigger for storms.

The storms in northeast OK and southeast KS formed during a window when it was still in the warm sector, with sufficient instability, the cold front had not arrived yet, but the upper-level trough had finally come close enough to provide support for storm formation. (I still suspect a gravity wave may have been involved because of the sudden, simultaneous way these storms exploded, but cannot prove or disprove this.)

As far as when storms die, I believe much has to do with storm mode. A bona fide supercell will have the regions of updraft and downdraft separated, and can stay alive for several hours. For ordinary thunderstorms, the downdraft will often fall down on the updraft and cut off it's own fuel. For forecast purposes, you can look at deep layer shear to help determine when conditions are favorable for supercell formation.

 

[Rick Schmidt]

Just on a quick note regarding supercell life, I believe the Plainville Kansas supercell in 1992 lasted for 10 hrs. Maybe not a record, but it's close. It also spawned around 12-14 tornadoes.

 

[Jeff Duda]

Clarifying Question: When you say, "mesoscale enhancement to the surface flow near the DL to the north since that area had been worked over by a morning MCS, thus providing enhanced lift to get storms to fire there" does that mean an outflow boundary?

Basically, yes. The outflow probably forced the winds to back (become more SEly), although I see now from looking at surface winds from around the time the storms developed little evidence to support that hypothesis. Perhaps the gravity wave hypothesis then provides more of the answer to this. Honestly I don't know anything about that, though.

Clarifying Question: So am I right to believe that as storms cross a warm front they actually have a small window where interacting with the warm front actually strengthens them, or enhances the chance of tornadogenesis? But, once they sufficiently pass the warm front, they die off rapidly?

Definitely. I have seen this more than a few times. I even documented such an incidence where a tornado formed close to home while I was out chasing elsewhere: http://www.meteor.iastate.edu/~jdduda/chasing/2009/042609.html

Big Question: What is the best way to determine if a storm will live or die. I have been wondering about this for a long time. Is it best to really study the atmosphere (with real time surface obs and models) the morning of the chase so that you know exactly what the atmosphere in front of the storms is supposed to be like and therefore you will know which storms to pick? Or is there a way to make educated guesses while you are actually chasing in order to know what storm to pick? Or, am I thinking too microscale on this, and once you have chosen a target area, it becomes a crap shoot and going tail-end-charlie is just the safest bet?

Fundamentally, a storm MUST ALWAYS be ingesting unstable parcels (i.e, must be in a region with non-zero CAPE). However, just having >0 CAPE does not guarantee a storm will continue. There must also not be too much CIN to keep parcels from reaching their LFC. Significant downward motion can also kill off a storm, although I don't believe forces on even the mesoscale can provide strong enough descent to do this, and I doubt such a factor ever causes a storm to die.

BTW, you can learn how the quoting works by clicking on the "reply with quote" button on someone's post and study the code that is automatically generated. It is quite simple code.

 

[David Williams]

On it again! Thanks Jeff. I see you have a few degrees in meteorology. I would love to hear what you have to say with regard to my question in the "advanced weather and chasing" section about tornado chasing and meteorology degrees.

BTW, you can learn how the quoting works by clicking on the "reply with quote" button on someone's post and study the code that is automatically generated. It is quite simple code.

 

[Matt Hunt]

Fundamentally, a storm MUST ALWAYS be ingesting unstable parcels (i.e, must be in a region with non-zero CAPE). However, just having >0 CAPE does not guarantee a storm will continue. There must also not be too much CIN to keep parcels from reaching their LFC. Significant downward motion can also kill off a storm, although I don't believe forces on even the mesoscale can provide strong enough descent to do this, and I doubt such a factor ever causes a storm to die.

David, this is basically what I looked at on 4/27 in western MO. After the QLCS went through, more storms fired, and even became severe warned early in their life cycle, but a quick check on the SPC Mesoanalysis page revealed there wasn't much CAPE in that area, and certainly not the area they were heading into. I do believe CIN was fairly high as well. So I passed on those storms (they did die off quickly) and headed toward Joplin where CAPE was already recovering, and forecast to increase through the evening, with no CIN. I ended up seeing the Baxter Springs tornado from the south side of Joplin.