I have often pondered this question about ENSO contributions to
interannual variations in global average temperature. I wonder if it
has been fully addressed in the literature, hence a reference to a
good paper on this would be helpful if someone has one.
Oliver mentions the ocean, and certainly this plays a part. In the
eastern tropical pacific, the typical condition has the oceanic
thermocline shallow enough that the trade winds driven upwelling draws
cool water from below the thermocline to the surface. During El Nino,
the thermocline dips and the cool sub-thermocline water is now too
deep for the upwelling to draw on. Thus, we typically think of "warm
water moving from the west pacific to the east pacific" during an El
Nino, but I think it is more accurate to say that the eastern pacific
warms while the west pacific stays about the same temperature, and the
warming of the east has more to do with vertical mixing in the ocean
than west-to-east mixing. So El Nino shuts down a process in the
eastern pacific that under non-El Nino conditions acts to cool the
atmosphere. This is certainly an important factor, but I've often
wondered if this is really sufficient to explain the global-scale warm
anomaly.
Oliver also mentions albedo and long-wave absorption. Given that
clouds impact both of these, and dramatic changes in tropical cloud
patterns accompany El Nino, I've always assumed that these changes in
clouds play a role in the global temperature anomaly. Alastair
mentions cloud-induced warming, but the cloud albedo contributes a
cooling that is comparable in magnitude to the warming from the cloud
greenhouse effect, so it may not be as simple as saying there are more
clouds. Someone must have looked at the changes in cloud radiative
forcing during El Nino. Anyone know of the paper?
Regarding "the amount of energy that can be released/soaked up by the
oceans is finite": that might be true, but we're a long way from
warming the deep ocean as much as it could be, and the time it takes
to mix thermal anomalies through the entire depth of the ocean is
millennial in scale. So by all means, changes in ocean currents that
alter the rate of vertical mixing of thermal energy in the ocean can
have profound impacts on the global average surface temperature.
-Eric