Hi all,
The very active April 2011 severe weather season across the United
States looks to finally be drawing to an end, capped off by yesterday's
historic tornado outbreak across the Deep South (for a radar loop, see
http://moe.met.fsu.edu/~acevans/cases/radar_animation.gif;
credit: College of DuPage and Dave Ahijevich, NCAR/MMM). There have
been no less than five significant severe events across the
nation this month: April 4, April 9-10, April 14-16, April 19, and April
25-27. A bit on the historic nature of this month's activity is
available at
http://www.npr.org/blogs/thetwo-way/2011/04/28/135806635/putting-the-months-severe-weather-outbreak-in-historical-perspective?sc=tw, though it should be noted that the count for this month is preliminary at this point in time.
With that in mind, I figured that it would be worth looking at the
synoptic-scale conditions across the region for April 2011 to attempt to
understand what factors, if any, potentially led to this clustering of
high impact events. I should note that my goal
is not to analyze each individual case to see why and how severe modes
evolved as they did; rather, I'm looking at the larger-scale throughout
much of this discussion. Before doing so, however, I feel that a hearty
thank you for a job well done is in order
to all of those on the list and elsewhere who were on the front lines of
everything yesterday. Watch/warning/dissemination operations seemed to
work very, very well yesterday and it's quite possible that the fatality
count would have been a lot higher were
that not the case.
That said, moving into the April 2011 analysis, consistent with a La
Nina-influenced seasonal weather pattern, we've seen rather persistent
subtropical ridging centered over the far southwestern North Atlantic
since February. This is reflected in a generally
positive NAO signature (CPC analysis attached) and well above normal
SSTs across the Gulf of Mexico (SST analysis and anomalies attached).
During the month of April, this pattern has become even more pronounced
as the polar vortex, or what passes for it this
time of year, has become quasi-stationary across northern and
northeastern Canada. To its south, we've seen persistent ridging for
over two weeks that has only slowly retrogressed westward in the last
week or so. This is highlighted well by the attached 500
hPa analysis and anomaly animation from the CPC that also implies a
slightly confluent upper level pattern across the western North
Atlantic.
Across the western and central United States, the pattern hasn't been
significantly amplified. We've had a variety of disturbances enter the
western via an at times energetic Pacific jet stream. Many of these
disturbances amplified slightly across the central
US, but there were a variety of ways in which they got to that point.
Heather Archambault's DT analyses from April 2011 are particularly
insightful in highlighting this and are available at
http://www.atmos.albany.edu/student/heathera/dt/nam/1_to_15_apr.html and
http://www.atmos.albany.edu/student/heathera/dt/nam/15_to_30_apr.html.
For instance, the most recent event appears to have had origins in a
split flow/quasi-blocking regime
across the western United States. An upper low entered the Gulf of
Alaska on 22-23 April and amplified the ridge across the western North
American coastline. A strong Pacific jet to its south, however, appeared
to undercut this ridge, resulting in a split
flow pattern by which upper level energy could crash into the western
US. The theme between events, however, is that the pattern has not been
overly amplified with any particularly event; in fact, most features
have lifted and/or deamplified as they have moved
eastward. The non-strongly perturbed flow has helped to result in an
anomalously westerly component to the flow aloft across much of the
nation this month (attached analysis from ESRL), a feature that is able
to result in enhanced directional vertical wind
shear assuming that the surface winds are not veered with each surface
cyclogenesis event.
This pattern, featuring a ridge over the subtropical Atlantic and a
continual train of shortwave troughs and accompanying lee-side
cyclogenesis on the High Plains, has led to anomalously strong southerly
flow into the south-central and southeast US over the
past month. This is well captured by the second panel of the CPC's storm
tracks & 925 hPa winds (vectors) and wind anomalies (shaded)
analysis (see attached). Note that this signal has been particularly
strong over the past 10 days (first panel of the aforementioned
image). Of course, the signal is convoluted by the presence of the
cyclones, but even over a 90 day period (third panel), there is a
notable positive anomaly into the south-central US. This has two
effects. First, this helps to enhance low level vertical wind
shear (and presumably helicity) across the region. Deep layer vertical
wind shear, particularly of the directional variety, is also enhanced
(c.f. the 500 hPa zonal wind anomalies with the 925 hPa wind anomalies).
Secondly, this has helped to result in anomalously
high boundary layer moisture across the region during the month of
April, as shown by the attached reanalysis-based 1-26 April 2011 mean
925 hPa specific humidity field (attached analysis from ESRL; contour
interval: 0.5 g/kg). This has presumably been influenced
by both advective processes from the Gulf (with the above-normal SSTs)
as well as lifting associated with the low level jet. The lack of a high
amplitude wave train has likely also prevented Gulf moisture from being
scoured out by cP air masses at times during
the month; indeed, multiple days of moistening from the Gulf was
observed across the southern US ahead of the most recent severe event.
In the aggregate, we have a pattern favorable for repeated moderate to
high impact severe weather events across the nation throughout much of
April. A few questions arise from all of this, of course, related to
both science and societal issues.
1) What actually caused the aforementioned flow configurations? Is it
largely all a reflection of the (weakening) La Nina base state? What
else is at play?
2) How rare are these conditions over a month-long interval?
Standardized anomaly analyses of these and related fields would be
helpful in this regard. I suspect that the combination of energetic
low-amplitude flow coupled with rich boundary layer moisture
is rather rare for April and more typical of May and early June.
3) Focusing on yesterday's event in specific for a moment, the storm
that impacted both Tuscaloosa and Birmingham maintained itself as a
discrete supercell into the southern Appalachians with a number of
tornadoes likely to be surveyed along its path. We've
seen a number of long-lived supercells with various events in recent
years; apart from this specific cell, I can recall one on 2/17/08 in the
same geographic area as well as the Greensburg, KS supercell as two
other examples. What is unique about these supercells?
At a basic level, cold pools seem to be mitigated to some degree with
such storms and, indeed, it was fairly moist aloft and in the boundary
layer yesterday (see attached BMX sounding from SPC valid 28/00Z).
There's probably also a hydrometeor/microphysical-
related
component (e.g., James and Markowski 2010, MWR) to the cold pool
evolution as well. Where do the controls on singular discrete supercell
evolution lie: the synoptic-scale, mesoscale, microscale, or some
combination thereof? Bunkers et al. (2006a,b, both in
WAF) discuss the environments associated with long-lived supercells and
can likely shed some light on these questions. Finally, though
predictability of large-scale storm mode may be at least modestly
predictable in a forecastability sense, how likely are
we to be able to predict that an individual supercell will be of long
duration whereas another supercell in at least a somewhat similar
environment will not?
4) At this point, it looks like the final death toll will be in the
vicinity of 300. Even with this in mind, the event was very well
forecast at all lead times, both short and long. The SPC, impacted NWS
offices, and local media did a *great* job forecasting,
analyzing, and disseminating information yesterday. This begs the
question: are ~200-300 fatalities what we should expect from such a high
impact event? If so, why? To me, this is not so much of a science issue
as it is a socioeconomic and "individual response
to science/information" issue, yet still important given the growing
focus of funding agencies on the broader impacts of scientific research.
When getting at why ~300 deaths occurred with this event, a few
questions arise. Is it an infrastructure issue? If so, are community
shelters for mobile homes/apartments/businesses and an increased density
of basement facilities for single-family structures
a viable answer that can save lives? Is it a reactionary issue? I
suspect that a lot of people reacted in a reasonable way to the warning
information they received, particularly given the active cycle we've
been in this month, and I wonder if we will find
that despite doing so, the infrastructure in place was not sufficient to
save their lives. Is it a socioeconomic issue? How will people's
responses to similar but (hopefully) less significant events evolve in
the future?
I don't think we've at all failed as a scientific community to convey
the science for this case. I think we succeeded on all fronts with that,
actually. Yet, I admit to it being somewhat sobering to know that even
if we do a great job that there are other factors
that may still result in the loss of numerous human lives, never mind
the millions to billions in property damage.
Lots of questions to keep us all busy for a very long time, both in
discussion here on the list as well as in research/operations-mode.
-Clark
--
Clark Evans
Postdoctoral Fellow, ASP/MMM
National Center for Atmospheric Research
P.O. Box 3000, Boulder, CO 80307
E-mail: evans@ucar.edu
Phone: 303.497.8927
