Cyclone Workshop 1997 Meeting Summary
Tuesday 23 September 1997
authored by David Schultz

Session IV:  Richard Reed and Frederick Sanders Symposium
Chair:  Dan Keyser

The session was opened by Sanders' first Ph.D. student, Lance Bosart,
who recalled his thesis defense, attended by both Reed and Sanders who
argued with each other throughout his defense, thereby deflecting
interrogation away from the doctoral student.  Bosart was later told
that he was the first synoptic student to pass the M.I.T. Ph.D. exam
in meteorology.  In Bosart's eyes, Reed and Sanders' biggest
contribution was the high research standards that they set for future
generations.  John Gyakum continued with his recollections, in which
the meetings he often had with Sanders would generate many potential
Ph.D. theses.  Gyakum mentioned Sanders' extensive editorial efforts
including his often-lengthy reviews he would write himself.  Gyakum
noted that it was a tribute to their longevity that some of the most
insightful questions during yesterday's sessions were being asked by
Reed and Sanders.

The title of Richard Reed's presentation was "Perspectives on polar
lows gained from recent numerical experiments, or, Is there such as
thing as a polar low?"  As defined by Reed, the three characteristics
of a polar low is that it: (a) forms poleward of frontal zones and
jets, (b) lacks pronounced frontal structure throughout all or most of
its lifetime, and (c) appears as a comma-shaped cloud mass, in part of
convective origin.  Reed found it humorous that he was defining a
polar low, a phenomenon which he questions its existence.  He then
proceeded to describe the observations and mesoscale-model simulations
of three purported polar lows: 6-8 March 1977 in the Bering Sea, 26-27
February 1984 in the Norwegian Sea, and 23-27 January 1982 in the
Mediterranean Sea.  Model sensitivity studies indicated that while all
three lows needed latent heating to develop to their full intensity,
the Bering Sea cyclone did not develop at all without latent heating.
To answer the question in the title of his talk, Reed answered "yes
and no".  He claimed that there is a spectrum of events which depends
on the upper-level features, lower-tropospheric stability, latent
heating, surface fluxes, and baroclinicity.  Phil Smith expressed
concern about what time during the life cycle of the cyclone the
fluxes were turned off in the model simulations, a concern echoed by
Reed.  Steve Businger noted that the Bering Sea cyclone became more
symmetrical later in its life cycle.  Mohan Ramamurthy and Reed
debated whether polar lows formed over land in association with cold
vortices aloft.  Steven Mullen added that they exist, but they do not
reach the same intensity as their oceanic counterparts due to greater
surface friction and the lack of moisture.  Mel Shapiro and Paul
Neiman presented an analysis of observations from 1 November 1991 over
the western North Atlantic which had a 7 degrees C warm core.  Lou
Uccellini noted that a second system formed during that event and that
the phase relationships between the upper- and lower-level features
were similar to Deborah Hanley's research on tropical cyclones.  He
also noted that two out of the three cases had a double upper-level
jet structure (Kocin and Uccellini 1990, Fig.~21).

Fred Sanders presented "A method for analysis of surface maps."  The
inspiration for this research originated at the NMC Surface Analysis
Workshop (Uccellini et al. 1992).  The workshop led to a 34-page
letter which Sanders wrote to Lou Uccellini lamenting the state of
surface analysis at NMC.  The workshop also led to the research which
would result in Sanders and Doswell (1995).  These efforts emphasized
the importance of surface isotherm analysis which is not carried out
by NMC or by many researchers.  Sanders suspected that the Petterssen
(1940, p. 7) quote that "[surface temperature] is often neither
representative nor conservative" led to the demise of operational
surface isotherm analyses.  Sanders then proceeded to discuss the
conditions under which isotherms or isentropes should be analysed and
the conditions under which del_p theta approximates del_sfc T.
Sanders also considered moderate/strong surface temperature gradients
to be 8 degrees C per 220/110 km.  In response to concerns that the
fields were patchy and possibly unrelated to actual meteorological
phenomena, Sanders admonished that, "complexity should not be cause
for despair," and that these regions can be relevant to forecasters.
Sanders envisioned a spectrum of features which range from a
nonfrontal baroclinic zone (temperature gradient not accompanied by a
cyclonic windshift), to a frontal zone (temperature gradient
coincident with windshift), to a baroclinic trough (windshift without
a temperature gradient).  This classification scheme is consistent
with the Hoskins and Bretherton (1972) mechanism of frontal formation:
a gradient of temperature in a geostrophic confluent flow will result
in an ageostrophic circulation which intensifies the frontal zone.
Therefore, nonfrontal baroclinic zones are locations of possible
future frontogenesis.  Sanders pointed out the ironic statement often
made by forecasters during ERICA that the approaching "cold front"
would be associated with very little temperature gradient.  Sanders
used an example from his analytic cyclone model (Sanders 1971,
Fig. 10) to illustrate how the pressure trough and thermal gradient
could become separated.  Sanders recommended operational
implementation of manual or automated surface isothermal analysis, but
indicated that the analyst's interpretation would still take time.
Chris Davis indicated additional problems with selecting only
temperature, potential temperature, or equivalent potential
temperature and, in particular, rejected use of equivalent potential
temperature for measuring density contrasts, while recognizing its
importance for convective forecasting.  Reed and Keyser noted that
Milhelm (1910)? and Henry et al. (1916), respectively, performed both
thermal and baric analyses, but the Norwegian analysis methods may
have underemphasized the thermal.  Chester Newton noted that Milhelm's
book had so many observations that you couldn't see the isotherms.
Lou Uccellini noted that five positions at NCEP were eliminated
from the surface-analysis team and that the automated
approach will be inevitable.  He also suggested that the RUC is perhaps
the comprise solution.  Chuck Doswell said that when he was at SELS in
the 1970s, they performed hourly analyses of temperature and
dewpoints, whereas the local offices generally performed 
baric-only maps.  Consequently, "the final product ignores the
process by which it was derived."  Phil Smith wondered whether
Sanders' method would result in an increase in the number of short
frontal fragments, to which Sanders queried whether "the forecaster
would prefer one inch of front or six inches of lie?"

The final aspect of the tribute to Reed and Sanders was testimonials
from the audience.  Lou Uccellini identified Palmen and Newton (1969)
as the Bible for synoptic meteorology and anecdotally noted that the
definition of jet streak was found on page 199.  Newton was also on
Uccellini's Ph.D. committee and, as editor for Monthly Weather Review,
accepted Uccellini's first gravity-wave paper.  Uccellini recognized
Reed for his development equation based on the 500-hPa prognostic
charts and Sanders for documenting the operational models increasing
ability to forecast extratropical cyclogenesis (e.g., Sanders
19. . . .).  And as a forecaster at LaGuardia Airport, Sanders
successfully forecasted the 24-26 December 1947 snowstorm (Kocin and
Uccellini 1990, p. 9).  Steve Tracton recalled Sanders' nickname at
M.I.T., "old Dad," and told the tale of how he convinced Tracton to
pursue graduate school.  Cliff Mass humorously presented the surface
weather maps for the days that Sanders and Reed were born.  He noted
the passion for what they do and that they considered themselves
revolutionaries for critiquing the Norwegian cyclone model.  Chuck
Doswell recalled the time when Sanders, Harold Brooks, and he visited
Civil War Battlefields and that all three could get a senior discount,
if they were related.  So, now Sanders is affectionately known as
"Gramps."  Kerry Emanuel noted that he first became aware of Sanders
when his family got beat in a yacht race at Marblehead by Sanders'
boat.  Sanders later became Emanuel's freshman advisor at M.I.T.  He also
read a quote from Sanders' Ph.D. thesis regarding improper surface
analysis which could have been written last week.  Howie Bluestein
noted that Sanders' forecasting contest at M.I.T. was the model for
forecast contests at universities across the country.  Mel Shapiro
noted that his decision to pursue frontal research was inspired by an
FSU homework assignment to read Newton (1956?), Sanders (1955), and
Reed and Sanders (1953).  Other brief tributes were presented by Steve
Businger, Warren Blier, Chris Davis, John Nielsen-Gammon, Mankin Mak,
and Jon Martin.  Dan Keyser closed the session by highlighting the
mentorship that Reed and Sanders inspired him to attain.

Session V:  Paradigms of Cyclogenesis
Chair: Charles Doswell

The workshop continued after lunch with the lead-off paper presented
by Franc Ayrault (with Alain Joly and Beatrice Pouponneau) who
examined 850- and 300-hPa ECMWF analyses for secondary cyclogenesis.
He noted that low-level anomalies cannot initiate cyclones by
themselves in a statistically meaningful sense.  He also noted that
upper and lower anomalies do not necessarily grow simultaneously and
that a reactivation of an old system plays an important role in new
precursors, particularly upper-level anomalies.

Robert Cohen (with David Schultz) discussed the difference between
airmasses (separated by fronts) and airstreams (separated by airstream
boundaries).  He showed that the structure of airstreams depended upon
the large-scale flow in which the cyclone was embedded: meridionally
(zonally) oriented airstream boundaries were favored for cyclones
embedded in large-scale diffluence (confluence).  Reed noted that,
based on his experience, model simulations in which the latent heat is
turned off yielded clearer results.  Steve Colucci and Phil Smith
asked for more specifics on how Cohen's contraction rate is defined
and how it is related to an airstream boundary, respectively.

The next two talks (coauthored with Dan Keyser) dealt with the issue
of jet streaks as coherent disturbances on the tropopause.  Phil
Cunningham demonstrated that jet streaks could be modelled in a
nondivergent barotropic model as a superposition of dipolar vortices.
Despite the simplicity of their model, aspects such as the
propagation, structure, motion, and evolution of jet streaks could be
explained.  He stressed that the vorticity, not the wind field, was
more valuable in diagnosing these features and suggested that the
asymmetry of the vorticity field in observed jet streaks might be
explained by the presence of a background flow.  Chris Davis noted
that jet streaks appear on many scales and inquired whether
Cunningham's paradigm applied on all scales, to which Cunningham
replied that different scale jet streaks were apparent in his
simulations with background flows and were governed by the same
processes regardless of scale.  Matt Pyle then presented observations
of a nearly straight jet streak that resembled Shapiro's (1982) warm
advection case.  Pyle illustrated how the cyclonic anomaly could be
transferred from one jet streak in northwesterly flow to another in
southwesterly flow.  Mel Shapiro pondered how to reconcile Pyle's
balanced jet dynamics with Uccellini's unbalanced jet dynamics.  Davis
expressed concern that the QG and ECMWF omega were so disparate, a
topic that Pyle was to explore more in the future.  Mankin Mak
wondered how Farrell's optimal modes related to this work.

Mark Sinclair presented a composite analysis of extratropical
cyclogenesis in the Southern Hemisphere.  He identified four
categories which were related to previous work by Evans et al. (1994)
and Schultz et al. (1998): upstream jet-exit region cyclogenesis,
equatorward jet-entrance region cyclogenesis, downstream jet-exit
region cyclogenesis, and trough cyclogenesis.  Sinclair noted that in
the trough cyclogenesis category, the surface cyclone forms almost
directly beneath the upper-level vorticity maximum and may be unique
to the Southern Hemisphere.  Greg Hakim inquired about how long the
upper-level cyclogenetic precursors could be tracked.  Dan Keyser
encouraged Sinclair to continue his methodology for the Northern
Hemisphere and for tropical cyclogenesis.  John Nielsen-Gammon asked
if there were any classes where the upper-level precursors were
stronger than the low-level precursors, to which Sinclair replied that
the vorticity tendencies were indicative of such development.  Tim
Hewson wondered what the forcing for vertical motion looked like and
Sinclair noted that they could be related to the cloud patterns.

The session closed with Joe Trout (with Carl Kreitzberg) comparing
near-first-order discontinuities in temperature and wind observed
during aircraft frontal penetrations in ERICA to mesoscale model
simulations of those same fronts.  He concluded that the model
produced fronts that were too wide and weak compared to those
observed.


REFERENCES:

Evans et al. 1994

Henry et al. 1916

Hoskins and Bretherton 1972

Kocin and Uccellini 1990

Milhelm 1910

Newton 1956

Palmen and Newton 1969

Petterssen 1940

Reed and Sanders 1953

Sanders 1955

Sanders 1971

Sanders and Doswell 1995

Schultz et al. 1998

Shapiro 1982

Uccellini et al. 1992