TS Arlene -- another kind of circular MCS

Fittingly enough, as I began this discussion on mesoscale convective systems (MCSs), our first tropical storm of the Atlantic hurricane season formed off the east coast of Mexico.  Tropical Storm Arlene made landfall this morning near Cabo Rojo, Mexico.  Here's the an IR satellite image shortly after landfall:

GOES IR satellite image of TS Arlene at 1715Z, Jun 30, 2011.

I find it fascinating comparing the satellite presence of a tropical storm with the satellite presence of a mesoscale convective complex (MCC) (or a mesoscale convective vortex--but that's another post).  In Maddox (1980)'s classification scheme, the two types of recognized "circular" MCSs were tropical cyclones and MCCs.  As odd as it might sound, I believe that the origin of the "circular" structure is actually very similar in both cases, even if the both types of storms form in vastly different ways.

One main difference between circular MCSs (like tropical cyclones and MCCs) and more linear MCS features (like squall lines) is the environment in which they form.  Remember from my previous discussion that squall lines usually form along or near frontal boundaries.  Their structure and maintenence are dictated from the surrounding environment.  In fact, a lot of their energy is derived from a baroclinic environment.  That's a fancy way of describing an environment where there are strong horizontal temperature contrasts like we see around fronts.  But, remember that a MCC tends to form away from fronts--off by itself, more or less.  Tropical cyclones, too, don't have fronts in the picture--even if they are powerful low-pressure centers, they don't have fronts associated with them.  This is because both MCCs and tropical cyclones tend to form in a barotropic environment--an environment where there are not strong horizontal temperature contrasts.

This makes sense if you think about it in a qualitative sense.  Fronts are often depicted on weather maps as long, "straight" lines.  When we consider the polar front separating the cold, arctic air in the north from the warm, subtropical air to the south, it's a long, linear boundary that's asymmetrical on either side--one side is cold and the other side is warm.  Storms that form along these boundaries between warm and cold air (or moist and dry air) are going to inherit this asymmetry--the parts of the storms in the warmer, moister air are going to behave differently than the parts of the storms in the cooler, drier air.  This works against having a circular, symemtrical structure.

In contrast, storms that form away from strong frontal boundaries (i.e. in a barotropic environment) don't have these contrasting areas of warm and cold to deal with.  This is particularly true with hurricanes--they form over the tropical oceans where everything is very warm and very moist all the time (more or less).  As such, anything that spins (or has a vorticity maximum) in such an environment should (theoretically) assume a more symmetrical, circular shape.  If all parts of the storm are experiencing similar environments, they should theoretically all look the same.  Granted, tropical cyclones do often move into areas where there is differential wind shear aloft or cooling sea-surface temperatures and this disrupts that symmetry and weakens the storm.  But still--the circular structure makes more sense.

What about MCCs?  They're not usually associated with a powerful center of low-pressure.  However, remember what I said about MCCs in my last post--MCCs tend to modify their surrounding environment more than the surrounding environment drives them.  At the heart of any MCC is an area of rising motion and a lot of latent heat release as moisture evaporates when air is lifted and cooled.  That rising motion and heating lowers the pressure in the center, creating a warm-core-like vortex.  It may be weak and subtle, but it's the same kind of mechanism that drives tropical cyclone formation and growth.  And it implies that there still is some (however weak) vorticity and circulation going on in the storm.  Is this enough to really promote a circular structure?  That's my guess for now.  Though if anyone has a better idea, I'd welcome any thoughts...

Going back to TS Arlene, it was interesting to note that our global models were forecasting a tropical cyclone to develop in the western Gulf of Mexico a full week ago--an impressive lead time for tropical cyclone genesis.  Here's the ECMWF 144 hour forecast of sea-level pressure for 00Z yesterday:

Note that the model forecasted a small, compact low in the Bay of Campeche just off the eastern Mexican coast.  It's a little further south than where the storm actually impacted the coast, but still rather impressive with such a long lead time.

Here's the same forecast from the GFS model:

The forecast low-pressure center was a significantly more pronounced, though still a little far south.

And here's the CMC model (Canadian GEM):

The CMC model was predicting a much deeper low.  For some reason the Canadian model always seems to predict tropical cyclones that are much stronger and much more frequent than actually happens.  I'm not entirely sure why this is, but it's an interesting question.