Class Notes April 12 , 2001

Summary of Thursday April 12, 2001 ATM S 560 lecture							Sarah Gaichas

What is El Nino/Southern Oscillation (ENSO)?
--------------------------------------------
In this lecture, we reviewed the essential observations of ENSO, which no
model should be without.  However, the first caveat is the available data:
there are only about 50 years of SST data for the tropical Pacific, and much
of this was collected at only a few points along the equator in shipping
lanes.  There are observations of surface winds since about the 1950 s, and
the observations of vertical structure in the ocean go back only about 20-30
years.  This short time series makes it difficult to evaluate the power
spectra of the ENSO indices at greater than interannual timescales; decadal
scale signals are virtually impossible to resolve.  Fortunately, the
magnitude of the East West SST gradient in the tropical Pacific is large
enough to be reasonably well described even with few observations along the
equator.

Description of cycle 
--------------------
The ENSO phenomenon cannot be explained by the physics
of the atmosphere or the ocean alone, this is a coupled system in which
there is atmosphere to ocean influence via surface winds, and ocean to
atmosphere influence via SST anomaly-generated heat flux (both latent and
sensible).  ENSO is intrinsic to the tropical Pacific ocean, and its signal
propagates from the tropics to the mid latitudes according to observations.
During normal or (more extreme) La Nina conditions, the easterly surface
trade winds set up the elevated sea level and depressed thermocline in the
Western Pacific.  Therefore, while equatorial upwelling continues in the
western Pacific, it is of warmer water from above the thermocline, while
upwelled water in the eastern Pacific comes from below the thermocline and
is cooler.  This increases the zonal SST gradient, which in turn intensifies
the atmospheric pressure gradient, which increases the winds, sea level and
thermocline tilts, and therefore SST gradient in a positive feedback.  In
the ocean, horizontal advection is from cool eastern waters to warm western
waters (which may prevent the SST gradient from increasing forever?). The
zonal atmospheric circulation is driven by convection over the warm western
Pacific (with heavy precipitation there) with dry air sinking in the east at
the South American continent. These conditions are reversed during El Nino,
which is initiated by a relaxation of the easterly trades.  
                  (Why do the trades relax?)  
The relaxation of the trades permits a lower sea level and a
shoaling of the thermocline in the western Pacific, a disturbancewhich
flattens the thermocline across the ocean basin and raises sea level in the
eastern Pacific (via equatorial Kelvin waves?).  The deeper thermocline in
the east shuts off the source of cool water for upwelling, and upwelling is
weaker due to the weaker winds, so surface waters warm and the cross basin
SST gradient is reduced or eliminated.  The center of atmospheric convection
and the associated heavy precipitation is displaced to the central Pacific
over the warmest water, and surface winds are weakened further and may even
reverse and become westerly.  The ITCZ shifts towards the equator.  The El
Nino conditions are reversed with the strengthening of the easterly trades
(again, why do they come back?-I think we get to thatnext lecture).

Basic patterns in the observations
----------------------------------
The temporal and spatial characteristics of ENSO are reasonably well
described; changes in SST, sea level pressure, surface winds and heat flux
have known sign and amplitude within a factor of 2.  In general, warm (El
Nino) events are stronger than cool events (especially in the eastern
Pacific), but happen during about 1 year in the 3-7 year cycle.  The extreme
conditions do not occur at random times throughout the year; they are
coordinated by the seasonal cycle. During the 1980s peaks in SST anomalies
were often during northern hemisphere winter, but this has not always been
true; the time series is probably not long enough to determine whether the
seasonality observed during the 80s was typical or whether the non-winter
peaks observed before and since are the norm.  Most of the variance in ENSO
indices occurs in northern hemisphere winter, with the least variance
between March and May.  While it can be argued that the equatorial Pacific
system should be most unstable between June and December when winds are
strongest and temperature gradients are largest, we noted that the water is
warmest during the spring months, which leads to the largest heat fluxes and
therefore increased atmospheric instability.  Finally, we examined the
statistical properties of the ENSO index anomalies, and found that the
distributions are nearly Gaussian, especially in the central Pacific
(although there was increasing skewness approaching the eastern Pacific).
This implies that either the most important processes operating are linear,
and/or that the non-linearities in the system cancel perfectly.  However,
skewness and kurtosis would be better described (as of course would
everything else) with more observations.  In addition to the tropical
Pacific ENSO signal, there are atmospheric teleconnections to the global
tropics and the extratropics, with systematic anomalies observed in the mid
latitudes.  We noted that ENSO still explains little of the winter to winter
variance in mid latidudes however; for example this year the Pacific
Northwest should be having a very wet winter according to predictions based
on the ENSO cycle, and not the dry winter we are having.

Specifics on heat flux in the eastern Pacific
---------------------------------------------
All things being equal, the reduced trades in the eastern Pacific should
mean reduced heat flux as well; but heat flux is observed to increase
dramatically during a warm event, especially along the equator.  Heat flux
of the observed magnitude should be able to remove the excess heat in 2 to 3
months, but warm events last much longer, up to six months. This indicates
not a local forcing of the atmosphere on the ocean, but a concentration of
warmer water in the eastern tropical Pacific due to oceanic circulation that
maintains the heat which is out of balance with the atmosphere.  The
components of surface heat flux include increased solar insolation right at
the equator due possibly to deeper convection and therefore less cloudiness
in the area, latent heat flux out of the ocean in the same area, and
increased heat flux from the atmosphere into the ocean immediately to the
north, resulting in a wide band of warm water.

We finished with a slight digression, answering the following questions (and
I may not have this quite right ):  Why is the ITCZ where it is? What would
it take to move it and does this explain a wide equatorial band of apparent
primary productivity in the last interglacial (which was originally
explained as a wider region of equatorial upwelling, a phenomenon that does
not seem physically likely)?

The ITCZ s normal location is probably a factor of the relative position of
the coastline (slanted, not straight north south) and the location of
mountains, combined with different patterns of Ekman transport in each
hemisphere.  Offshore winds north of the equator result in Ekman transport
up the coast (warm southern water transported north), and therefore warming
to the north.  In contrast, winds are more along the coast in the southern
hemisphere due to the slant of the continents and the high mountains (?)
resulting in Ekaman transport away from the coast, and the cooling of
offshore waters (cold deep and more southern water moving offshore).  The
end result is an atmospheric high pressure over the south Pacific
contrasting with relatively lower pressure just north of the Equator, and a
convergence zone there (the ITCZ).  If during the last glacial time the
maximum summer insolation was in the southern hemisphere rather than the
northern (as at present), this might reduce the contrast between the two
hemispheres by reducing the strength of the southern hemisphere high
pressure system, thus allowing the convergence to move further south,
possibly below the equator.  This would reverse the pattern of upwelling so
that it would then fall north of the equator, because the prevailing winds
would be towards the south across the equator.