Warm fronts and the snow pack

[Skip Talbot]

Hopefully somebody who knows a bit more about meteorology can set me straight on this one as I feel like a newb on this topic. I was cruising the models two nights ago trying to pick targets for yesterday's chase. There was a sharp temperature gradient across central IL, but not associated with a wind shift like frontal boundaries typically possess. I didn't immediately realize it and somebody had to point out to me that the snow pack was causing the sharp temperature gradient. I guess I'm not used to dealing with large areas of snow when making a chase forecast.

Lots of folks were still calling this gradient a warm front, however. Is this correct? It didn't seem so to me, especially without the wind shift. It seems like the warm airmass kept heading north with the warm air advection and that only a thin layer near the ground was being cooled. There was a wind shift well north of the snow line, up by Kankakee, IL by early afternoon although the temperature gradient was much more diffuse up there then it was further south by the snow line. If I had to label it, this is where I would put the warm front on a surface chart.

What about this snow pack cooling the surface layer though? Obviously it's an instability killer for surface based storms, but does it also act like a stationary frontal boundary? Even though the models aren't showing it, do the winds back any north of the gradient? Do we get any sort of favorable directional shear enhancement near the snow line? It seemed that several of the tornadoes that occurred yesterday happened immediately south of the snow line as if this was indeed acting like a warm front, but perhaps this was just coincidence as the parameters were maximized there at the time.

 

[Dan Robinson]

I had the exact same thought Skip. I expected the warm front to lift farther north, but it appeared to simply stall, which surprised me a little. From other chasers' lightning/storm images further north from the previous day/overnight, it was evident there was a more solid snowpack up there. I was actually concerned about the front moving into the solid snowpack zone, and losing the more ideal low-level CAPE in the lowest 1km due to the higher albedo.

I'm not sure how much the snowpack eroded yesterday, but the vis sat loop shows where it is now:

http://climate.cod.edu/flanis/satellite/1km/index.php?type=Missouri-vis-12

The surface obs did seem to indicate a classic warm frontal zone, with a sharp transition to easterlies to its north.

 

[chrisbray]

Great points Skip. I live in the Kankakee area, and we did not warm up as much as forecasted. I blame snowpack.

 

[James Gustina]

From what I saw, it looked like a baroclinic zone in and along that gradient/front thingy might have been what did it. The NWS definition has this:
Warm Front
A transition zone between a mass of warm air and the colder air it is replacing.

That makes me think it technically is a front even without the wind shift, but I'm not sure if it was actually replacing the airmass to the north or if it was just a sharp gradient that remained near stationary.

 

[Rob H]

From my understanding, a front is always co-located with a pressure trough - just a baroclinic zone isn't enough to define one. So that gradient would not be a warm front. The actual warm front made it into the snowpack area, and the model (and assuredly sfc obs) showed a tempered warming in the transition zone. Since there was no wind shift between the snowpack and not-snowpack areas, my hypothesis is that the snowpack's effects were enough to affect 2m temperatures by ~10deg, but not enough to create a pressure trough or anything else meteorologically significant, and therefore no wind shifts or enhanced directional shear.

So could the snowpack line affect tornadogenesis? Well there's been research and conjecture that links baroclinic zones caused by anvil shadows to enhanced tornadogenesis, so I'd imagine that it could be a factor here as well. Probably not a lot of historical cases to study, however.

Disclaimer: all of my knowledge comes from reading web sites, a few books, and black magic. I ain't seen no schoolin' for weather.

 

[Dan Robinson]

I'll have to go back and pull some data and look at this closer. I was having data issues during the chase, and was relying on ThreatNet for everything. I can't remember the ASOS stations off hand, but the ones north of I-72 were showing 36/36 and fog with easterly winds, and the ones to the south of I-72 were south-southeasterly and clear with dews into the 60s. I could see the telltale low clouds that usually mark the cold side of a sharp boundary a few miles north of I-72. I saw a lot of ground fog where there were larger areas of unmelted snow. Will be interesting to see if the snowpack created a secondary boundary.

 

[Stan Rose]

I wouldn't get overly hung up on the semantics or terminology, Skip. The real world rarely resembles the textbook definitions--just come on over to CO where all sorts of boundaries are called fronts whether or not they strictly agree with the 'official' definitions. It's pretty clear from the overall synoptic maps that there was a warm front over central IL at that time, and id put it where you would--coincident with the wind shift and isobar kink. Youd have to look at a nicely resolved cross section to see how the vertical temp gradient was setting up in that frontal zone. The surface characteristics may have had an effect on frontogenesis. If i were anxious to get my Phd i might look more closely, but i think im gonna get an MD first, it will look cooler after my name.

 

[Adam Lucio]

It took longer than forecast, but the temperature at Midway Airport (chicago) near my house jumped from 39 to 52 in less than 30 minutes. When it did the visibility went from near zero to crystal clear. It was interesting...but that snowpack definitely hindered the northward advancement of it. Otherwise the severe weather would heave easily spread into Northern Illinois.

But then again I think about how if we didn't have such a large and expansive snowpack with re-enforced cold air like that, the low wouldnt have deepened the way it did anyways. Ah weather...

 

[Mike Johnston]

Sounds like the SPC did a good job of taking the snowpack into account for this event. From the 2/20 morning convective outlook:

"LOSS OF SFC-BASED PARCELS IS EXPECTED IN RELATION TO
PROGRESSIVELY DEEPER STABLE LAYER WITH NWD EXTENT...ATOP SNOW PACK."

 

[calvinkaskey]

You are right Stan. Things in text books are often too simplified. I think that small funnel clouds could be the norm along gust fronts. There should be something in textbooks and especially spotter training about them. I'm certain that with the extreme shear and instability along gust fronts that those 5-10 second long objects that are darker than surrounding clouds, connected to clouds and funnel shaped and don't move like scud are most definitely funnels. They are in places with the highest instability and thickest gust fronts.

 

[Jeff Duda]

I wouldn't get overly hung up on the semantics or terminology, Skip. The real world rarely resembles the textbook definitions--just come on over to CO where all sorts of boundaries are called fronts whether or not they strictly agree with the 'official' definitions. It's pretty clear from the overall synoptic maps that there was a warm front over central IL at that time, and id put it where you would--coincident with the wind shift and isobar kink. Youd have to look at a nicely resolved cross section to see how the vertical temp gradient was setting up in that frontal zone. The surface characteristics may have had an effect on frontogenesis. If i were anxious to get my Phd i might look more closely, but i think im gonna get an MD first, it will look cooler after my name.

Agreed. Clearly you see it now and it sounds like your reasoning and analysis are good. A front is simply a boundary between two air masses - there is no requirement of a pressure trough or wind shift for a boundary to be called a front. It's just very common for such features to be associated with one given the frontal circulation.

Looking back at vis satellite from the time of the HRRR forecast you showed above, it looks like the model had the snow pack boundary a tad too far south.

On a more general note, models try to account for surface snow cover, but there are frequently errors associated with the snow pack analysis, which can cause models to be very wrong near the edges of snow packs. Typically that won't impact a forecast for severe weather since it's uncommon to have severe weather occur so close to a cold air mass with snow on the ground like that, especially once we get into April and beyond.

Also, I think if you could see a vertical profile above the snow pack but south of the warm front, you'd probably see an extremely shallow layer of snow-cooled air with a big warm nose over the top of it. Clearly, an ML parcel would not be the most representative for analyzing storm mode and strength. The MU parcel would probably do a much better job approximating the true storm environment. I don't think you could get a tornado in a situation like this (north of the gradient) because surface parcels would simply be too heavy. I suppose, however, if there was the mother of all warm noses above that, and the warm nose was saturated with an EML over the top such that you had 3000+ MUCAPE and a well-established storm, there could be enough momentum to pull up some parcels just after the storm crossed over the snow pack boundary and put out a quick tornado. It certainly would be quick, though.

Cool find.

 

[James Gustina]

The relative buoyancy of parcels immediately north of that zone was probably ridiculously negative, hence why it's all the more impressive that so many tornadoes got spat out so close to a stable airmass that was likely overrunning the warm sector. The stable layer must not have been too deep like Jeff said, just based off of what the storms looked like, with relatively low bases under 1250 m.

 

[Dan Robinson]

Here's a surface data plot at 20:07z:

Skip, do you remember if there was a snowpack in the morning up there along I-72? I figured it had retreated far to the north by morning, given the overnight/morning thunderstorms, and that only the deeper drifts remained in the I-72 corridor at 2PM. Due to that data hole between I-72 and Macomb, I was guessing on the location of the front at the time, assuming it was farther south due to the fog/low clouds visible to the north of I-72. That fog may very well have been due to the snowpack. Very interesting that these could have been "snowpack boundary" tornadoes.

 

[Rob H]

A front is simply a boundary between two air masses - there is no requirement of a pressure trough or wind shift for a boundary to be called a front. It's just very common for such features to be associated with one given the frontal circulation.

For my own knowledge, I've been told that a front always lies in a pressure trough. Margules' Frontal Slope Equation and relative minimums and all that. Is this not correct?

 

[Rob H]

For me at least, calling something that isn't a front a front is problematic. If there is a front I assume a whole suite of things - that a cyclone is involved, the air masses are not superficially deep, that there might be a wind shift with enhanced helicity, and most importantly that it's associated with pseudo-geostrophic processes. There's a big difference between all of that and differential heating from anvil shadows or cirrus holes, geographic features like the Buffalo Ridge or eastern edge of the caprock, snowpack, outflow etc.

Maybe I'm being too technical, but paying close attention to the nature of boundary anomalies feels like it has helped me be a more successful chaser.