4/1/2006 DISC: TX/OK/KS

OK, so I'll largely consider today a bust. At least it wasn't a clear-sky bust, but it certainly was a bust in terms of the number of supercells and the lack of tornadoes. Here are my 2 cents...

All models had forecast Tds in the 63-68 range across southwestern OK by afternoon. When I looked at the 3pm obs, it was obvious that the models were way overdoing surface moisture. It's interesting to note that the same was true of 3/12 for the most part. When I looked at the FWD sounding and saw the 140mb deep moist layer, and noticing that shear wasn't terribly strong, I figured we wouldn't need to worry about mixing out the good juice. However, OK mesonet obs didn't really show any sites >63F dewpoint west or north of I44 by afternoon (and all areas from the 1st-tier of counties in western OK and westward into the TX panhandle were <58F td). So, it was quite obvious that the real nice dewpoints stayed down in southcentral OK (and southeastward from there), where places like Ardmore and Duncan saw 65-69F dewpoints most of the day. So, it appears that (a) the richer moisture couldn't advect far enough westward (into western OK and eastern TX panhandle), and ( B) the moisture that was there mixed out a bit. This led to much greater dewpoint depressions (and much lower boundary-layer RH) than were (was) forecast. So, in any particular storm's downdraft, there was likely plenty of evaporational cooling (with low RH present), which in turn likely lead to stronger cold pools. With strong cold pools, more convection was forced along its leading edge. Eventually, we saw rather quick upscale growth to a linear mode. With modest upper-level flow, along with plenty of potential for cold downdrafts, I feel the end was result were outflow dominant storms.

Now, I think we may have actually had too much insolation today. As strong diabatic heating occurred, the mixing depth increased (boundary layer deepened), which helped mix out some of the higher across western OK (dewpoints in many southwestern and west-central OK mesonet sites dropped 2-4F in Td between 1pm and 5pm). This likely allowed a very large area to near the convective temperature. With lowered Tds, and more potential for evaporative cooling in the low-levels, strong downdrafts developed. Since there was very little capping left over a relatively large area, the convergence along the leading edge of the downdrafts / cold pools was likely sufficient to initiate more convection. In addition, the RFD was likely strongly negatively buoyant (in fact, an OUN warning decision update even addressed the low theta-e air in the RFD). So, add all of this together, and I think that may explain why we saw a rapid evolution to a linear mode. Perhaps if we had a little less insolation, a little stronger capping may have persisted, and this issue may not have been as dominant.

It's also worth noting that low-level shear wasn't particularly strong. FWIW, RUC/SPC mesoanalysis didn't show >250m2/s2 0-3km SRH until about 2z. The relatively weak low-level shear in extreme western OK and the eastern TX panhandle, may have allowed the cold pools to spread away from the storm more than desired. The mid-level was actually stronger than the models had forecast (by about 10-15kts in most places of the target area), with 60kts at AMA and 35kts at OUN. I'm going to chalk this up almost entirely to the low-level thermodynamic and kinematic fields (moreso the thermodynamic environment, however).

EDIT: I continue to see some sometimes-large discrepancies in the dewpoints reported at ASOS obs and those reported at OK Mesonet sites. For example, Vernon was reporting a 70F dewpoin the late afternoon. Just a few 10s of miles away, Tipton and Grandfield Mesonet sites were only reporting 61-62F dewpoints. On the whole, it seems as the ASOS obs tend to be 1-5F higher in dewpoint than nearby Mesonet obs. Now, the ASOS sites directly measure dewpoint (with a chilled mirror), while the Mesonet sites calculate dewpoint from temperature and RH. So, I'm not entirely sure which one to believe, but there seems to be better spatial continuity with Mesonet obs compared to ASOS obs.
 
So, in any particular storm's downdraft, there was likely plenty of evaporational cooling (with low RH present), which in turn likely lead to stronger cold pools. With strong cold pools, more convection was forced along its leading edge. Eventually, we saw rather quick upscale growth to a linear mode. With modest upper-level flow, along with plenty of potential for cold downdrafts, I feel the end was result were outflow dominant storms.

Now, I think we may have actually had too much insolation today. As strong diabatic heating occurred, the mixing depth increased (boundary layer deepened), which helped mix out some of the higher across western OK.[/b]

That is an excellent assement Jeff. The cold RFDs are a killer in setups where buoyancy is limited. Ahh, the wonderful world of Markowski! :)

It seems the S. Beckham County storm would have had a chance if not for seeding from the evolving MCS to the SW. A precip-loaded RFD, likely much colder than the ambient environment did in the soon-to-be-gorgeous classic cyclic tornadic supercell ;) The mixing of lower Tds and subsequent higher LCLs didn't do the situation much good, and thats where the isolation gets ya. Also, the dryline really never mixed out of the central TX Panhandle, and storms had to travel quite a long distance to get into a favorable thermodynamic area...but they never made it, because the cold pools got their party together too quickly. April Fools! :)
 
Now, I think we may have actually had too much insolation today. As strong diabatic heating occurred, the mixing depth increased (boundary layer deepened), which helped mix out some of the higher across western OK.

[/b]

Amen to that, I think this happens more often than not on many chase days. And not only does strong mixing kill surface moisture, raise the LCL heights, and destroy CAPE, it also eliminates the low-level shear by generating a more uniform profile of U wrt height. I am not a big fan of days requiring strong insolation, I'd much rather have low CINH with a cirrus and/or stratus deck and very low LCL heights, with maybe a county or two width of sunshine near the dryline to get storms started. In my opinion, a decoupled, stable (with respect to theta, though not theta-e) boundary layer in the warm sector is a necessary (but not sufficient) condition for widespread tornadic activity.
 
Now, I think we may have actually had too much insolation today. As strong diabatic heating occurred, the mixing depth increased (boundary layer deepened), which helped mix out some of the higher across western OK (dewpoints in many southwestern and west-central OK mesonet sites dropped 2-4F in Td between 1pm and 5pm).
[/b]
As much as it hurts to say, I agree with that. I don't think we get to many situations were this occurs. I'm also thinking that the long drought that has occured all winter into spring had a effect on 4/1 as well as other days. With the sfc being so dry, it helps to mix out the higher Td's. I dunno though, just a thought.
 
EDIT: I continue to see some sometimes-large discrepancies in the dewpoints reported at ASOS obs and those reported at OK Mesonet sites. For example, Vernon was reporting a 70F dewpoin the late afternoon. Just a few 10s of miles away, Tipton and Grandfield Mesonet sites were only reporting 61-62F dewpoints. On the whole, it seems as the ASOS obs tend to be 1-5F higher in dewpoint than nearby Mesonet obs. Now, the ASOS sites directly measure dewpoint (with a chilled mirror), while the Mesonet sites calculate dewpoint from temperature and RH. So, I'm not entirely sure which one to believe, but there seems to be better spatial continuity with Mesonet obs compared to ASOS obs.[/b]

I'm not sure about the reason and who is right. I've learned that the Mesonet RH/Temp sensors are not aspirated (and the rain gauges not heated in winter, so precip there sometimes is not being measured when it occurs, more when it melts) which could lead to an underestimating of humidities and overestimating of temperatures in light to moderate wind conditions due to radiation errors. I'll visit the NWS and mesonet stations of Watonga with a sling psychrometer to have an idea what's going on. I spent a long time to calibrate my weather station's sensor which led to the Mid 60s yesterday 12 miles east of Watonga.

Graph last 36 hours
 
I think the cold downdraft was a problem, but I wonder how much forcing the (ostensibly) weak low-level shear/high dewpoint depressions were responsible for. There was a fairly well-established storm complex to the southwest of the Elk City storm, so it is quite possible that ice crystals from the (anvil level) outflow of the Childress complex prodded the storm to HP. I noticed that there were little "showers" forming in the inflow region of the supercell which probably cooled the surface inflow a bit (which would have reduced instability and allowed the cold pool to advance much further than normal). I don't really think that these showers formed as a result of a weak cap because other towers to the southeast of the Elk City storm died in a hurry after being impacted by the apparent subsidence surrounding the maturing supercell. Just a thought...

Gabe
 
EDIT: I continue to see some sometimes-large discrepancies in the dewpoints reported at ASOS obs and those reported at OK Mesonet sites. For example, Vernon was reporting a 70F dewpoin the late afternoon. Just a few 10s of miles away, Tipton and Grandfield Mesonet sites were only reporting 61-62F dewpoints. On the whole, it seems as the ASOS obs tend to be 1-5F higher in dewpoint than nearby Mesonet obs. Now, the ASOS sites directly measure dewpoint (with a chilled mirror), while the Mesonet sites calculate dewpoint from temperature and RH. So, I'm not entirely sure which one to believe, but there seems to be better spatial continuity with Mesonet obs compared to ASOS obs.
[/b]

I've observed a high dewpoint bias with the FAA maintained ASOS sites over Oklahoma for many years. They typically run 2-6F too high. The NWS maintained sites are much better. Also, I'd certainly trust the Oklahoma (and West Texas) Mesonet dewpoint obs over the FAA maintained ASOS/AWOS dewpoints. These mesonet sites are properly sited, constructed, and serviced by meteorologists who really care about data quality. The original OK Mesonet instruments were chosen by my Meteorological Instruments professor at the University of Oklahoma (OU) back in the late '80's and he chose well. My chase partner (RJ Evans) collects mobile mesonet obs with professional grade RM Young equipment and his dewpoint obs closely match the Oklahoma Mesonet obs; the FAA ASOS dewpoints are typically too high.

I'd bet many/most of the FAA ASOS dewpoint obs are flawed in other states as well, but the Oklahoma mesonet just allows us to see the discrepancy. You'd think the FAA would try to correct this *systematic bias*, but they don't appear to care and will tell you the obs are within acceptable tollerance. :rolleyes:
 
OK, I have been waiting for somebody else to bring this up, but that's not going to happen so I will. Surely somebody else noticed how far ahead of the dryline the Elk City storm went up? The other storms that went up along the dryline were a solid 40 miles West of the Elk City storm. It was hard to pin down the dryline on Saturday because of the wide moisture gradient and I don't remember whether or not there was a well defined wind shift, but I was paying very close attention to a subtle moisture tounge that ran along the TX-OK border and the West edge of this is where the Elk City storm went up. There was also a storm SW of Woodward that started to go up along the same boundary, but I didn't pay attention to what ended up happening to it. What was the focus for convection? There wasn't a dryline bulge (although this area was very close to where the mid level jet streak was nosing in). Outflow boundaries are out of the question. The only thing I came up with was that it was a secondary dryline that the Elk City storm went up on. On visible satellite there was a very noticeable North-South boundary along the border, but it wasn't visible until right before the Elk City storm fired(it was noticeable on surface charts, but it looked like it hadn't made it as far North as Woodward though, I think there is a gap in the sensors there so I probably just missed it). I was watching the TX panhandle mesonet very closely all day so I am pretty sure I am right on this deal, but I just wanted to know if anyone else noticed this and what is your thinking on the matter. Was there just enough lift generated by the density differences along this subtle moisture gradient for storms to go up?

BTW...
The only other time there was a secondary dryline (that I can think of off the top of my head) was on May 29, 2004.
The storm going up so far East was one of the main reasons I thought the Elk City storm was hands down going to be the storm of the day. I thought that if this storm had the advantage of already being in the better moisture there was no way it wouldn't be the storm of the day. It's like getting a head start in a race. This is why I was so shocked at the time when it didn't go tornadic.
 
OK, I have been waiting for somebody else to bring this up, but that's not going to happen so I will. Surely somebody else noticed how far ahead of the dryline the Elk City storm went up? The other storms that went up along the dryline were a solid 40 miles West of the Elk City storm. It was hard to pin down the dryline on Saturday because of the wide moisture gradient and I don't remember whether or not there was a well defined wind shift, but I was paying very close attention to a subtle moisture tounge that ran along the TX-OK border and the West edge of this is where the Elk City storm went up. There was also a storm SW of Woodward that started to go up along the same boundary, but I didn't pay attention to what ended up happening to it. What was the focus for convection? There wasn't a dryline bulge (although this area was very close to where the mid level jet streak was nosing in). Outflow boundaries are out of the question. The only thing I came up with was that it was a secondary dryline that the Elk City storm went up on. On visible satellite there was a very noticeable North-South boundary along the border, but it wasn't visible until right before the Elk City storm fired(it was noticeable on surface charts, but it looked like it hadn't made it as far North as Woodward though, I think there is a gap in the sensors there so I probably just missed it). I was watching the TX panhandle mesonet very closely all day so I am pretty sure I am right on this deal, but I just wanted to know if anyone else noticed this and what is your thinking on the matter. Was there just enough lift generated by the density differences along this subtle moisture gradient for storms to go up?

BTW...
The only other time there was a secondary dryline (that I can think of off the top of my head) was on May 29, 2004.
The storm going up so far East was one of the main reasons I thought the Elk City storm was hands down going to be the storm of the day. I thought that if this storm had the advantage of already being in the better moisture there was no way it wouldn't be the storm of the day. It's like getting a head start in a race. This is why I was so shocked at the time when it didn't go tornadic. [/b]

Mike,

Double-dryline structures aren't entirely uncommon. 5-3-99 and 6-12-04 are two that come to my mind.

There was a surface boundary apparent apparent on satellite in far southwestern OK. If you look at Mesonet obs, the winds in Hollis (and, I think, the ob immediately north of that one) often had a slight westerly component (like SSW-S), while winds to the immediate east had a slight easterly component. This was a pretty small-scale boundary, winds obs to the west of Hollis gradually went back to having an easterly component. In addition, notice that almost every storm developed along the leading edge of enhanced cirrostratus / cirrus that was overspreading the Texas panhandle (well forecast by the operational models, IIRC). I wish I had saved a satellite image at the time, but you can see storms explode as the leading edge of this cloud-cover approached. I would bet that there was some enhanced vertical motion (aloft) along this leading edge, providing just enough uwpard motion to completely remove the cap and allow for whatever very small subtle boundary was present to initiate convection. Just a guess... As you noted, the primary dryline was considerably farther west into the central TX panhandle (just east of AMA). I do believe that the 15z or 18z RUC correctly forecast qpf in western OK by 0z.
 
Back
Top