Recent El Nino Southern Oscillation
When the Christ child comes a year late
Although the name El Niño (`little
boy/Christ child’) actually refers to the fact that most ENSO (El Niño
Southern Oscillation) events peak around Christmas, it also nicely
captures a troublesome property of the phenomenon: It behaves as
quirkily and unpredictably as a three-year-old child. Next winter’s El
Niño can be reasonably well predicted in summer (which basically amounts
to diagnosing a developing event), but for longer lead times
predictability drops quickly: the `spring predictability barrier’.
The basic physics of El Niño are well
understood (see figure): under normal (non-ENSO) conditions, the
easterly trade winds over the Pacific push warm water to Indonesia. Near
Peru this leads to upwelling of cold deep water. If the easterlies
weaken, the warm water `sloshes back’ eastwards as an equatorial Kelvin
wave. As a result, the water near coastal Peru gets warmer and heats the
overlying atmosphere. This makes the air rise locally and the westerly
wind anomaly over the Pacific is further amplified: a positive feedback
(Bjerknes feedback) that eventually leads to a fully developed El Niño.
But how does this process start?
IMAU director and oceanographer Will de
Ruijter discovered in observational data that El Niño events tend to be
preceded by cool anomalies in the Indian Ocean, more specifically over
the Seychelles Dome (SD) region northeast of Madagascar. After the cool
SD event in summer 2013 he predicted an El Niño for 2014/15. Meanwhile I
spent the first year of my PhD on data analysis to prove that the
relation between the SD and El Niño is indeed statistically significant.
Observations support two ways in which the West Indian Ocean might
influence El Niño (see figures a-d):
Figure:
Illustration of the mechanisms by which a cool Seychelles Dome (SD) /
SouthWest Indian Ocean in summer of year 0 supports the formation of an
El Niño after 1.5 years (winter year 1/2).
1) The cool SD cools the overlying air,
which then sinks. To compensate, the air over Indonesia rises (a). As
Indonesia is warm and humid, this rising might be nonlinearly amplified
so strongly so as to attract an (easterly) inflow over the West Pacific.
This will lead to a greater-than-normal warm water reservoir, ready to
be pushed over to the East Pacific (b).
2) A cool SD in summer tends to be
followed by a stronger intraseasonal (time scales of several weeks) wind
variability over the West Pacific in the next winter-spring (c). This
can trigger a first warm Kelvin wave, especially since there is a large
warm water reservoir in the west Pacific. Now the Bjerknes feedback
kicks in and a fully-grown El Niño develops (d).
This is a nice theory, and if it is
correct, it may help to circumnavigate the spring predictability
barrier. But unfortunately there was no El Niño 2014/15. Does this
suggest that we are wrong? Not necessarily. In spring 2014, the warm
water volume was about as large as in the period preceding the very
strong El Niño event in 1997/98. The wind variability was high, too. So
one could argue that the Indian Ocean has done its job, but that the
Bjerknes feedback was not activated – i.e. the atmosphere did not
respond to the initial East Pacific warming after the first Kelvin
waves. At least not immediately. Instead, the warm water lingered
through early 2015 and a fresh spell of intra-seasonal wind variability
triggered Kelvin waves that led to one of the strongest El Niño events
on record – the Christ child came a year late
.
Why did the Pacific give such a strong
response in 2015 while remaining passive after a similarly strong
forcing in 2014? We do not know – probably the whole oceanography
community was puzzled by the failure of El Niño to develop in 2014/15.
Maybe Qingyi Feng and Henk Dijkstra can shed some light on this
question. They are working on detecting the stability of the El Niño
system, using correlation-based networks.
Claudia Wieners, PhD student in the Ocean and Climate research group at IMAU, Utrecht University
http://news.imau.nl/?p=2641
Comments