30
28
26
24
22
30
28
26
24
22
20
.. ..
, ....
i~
..:~
r~
~. ..
.
'P.:.
.~
:=
~.~ . ~ .
.: ~ ..
7
,~
t'
~~
.~
~
;~
..
.
.. . ~:~
.. .;:~
...
i::. .~
.. """': '!"r T'
Atlantic
~,. .
'P!:.
~
." ~~
jl ~
PacifIC
~
~~ ~:.:
.
.
~
I:;;
r.
i~
.. ...~ I~ ~
.. • I-!"' rr
i"'
. .
Mar
Jun
Sep
Dec
Mar
Jun
Sep
Dec
Figure 3: Scatter plot of monthly mean sea surface temperature (OC) in the equatorial
cold tongue regions of the Atlantic and Pacific for individual year/months, grouped by
calendar month. The dots for each calendar month have been scattered along the x axis
to make them more visible and the calendar year is repeated. Based on data from the
Comprehensive Ocean Atmosphere Data Set (COADS) for 1946-85. From Mitchell and
Wallace (1992).
power of harmonic analysis for detecting weak periodic signals.
2.2 Quasi-periodic phenomena
Because climate research tends to be driven, in large part, by the quest
for prediction, quasi-periodic phenomena have attracted a disproportionate
amount of attention per unit explained variance in the climate record. The
literature of the first half of this century abounds with claims of periodicities and quasi-periodicities in the climate record; many of them related
to the ll-year cycle in solar activity and its harmonics and subharmonics.
Few, if any of these claims have withstood the test oftime. Decidedly fewer
quasi-periodicities have been reported since the advent of power spectrum
analysis, which imposes more rigorous standards for assessing statistical
significance.
By far the most notable, and perhaps the only universally recognized
quasi-periodic phenomenon in the climate system on the interannual to
interdecadal time scale is the regular downward progression of alternating,
zonally symmetric easterly and westerly wind regimes in the equatorial
28
26
24
22
30
28
26
24
22
20
.. ..
, ....
i~
..:~
r~
~. ..
.
'P.:.
.~
:=
~.~ . ~ .
.: ~ ..
7
,~
t'
~~
.~
~
;~
..
.
.. . ~:~
.. .;:~
...
i::. .~
.. """': '!"r T'
Atlantic
~,. .
'P!:.
~
." ~~
jl ~
PacifIC
~
~~ ~:.:
.
.
~
I:;;
r.
i~
.. ...~ I~ ~
.. • I-!"' rr
i"'
. .
Mar
Jun
Sep
Dec
Mar
Jun
Sep
Dec
Figure 3: Scatter plot of monthly mean sea surface temperature (OC) in the equatorial
cold tongue regions of the Atlantic and Pacific for individual year/months, grouped by
calendar month. The dots for each calendar month have been scattered along the x axis
to make them more visible and the calendar year is repeated. Based on data from the
Comprehensive Ocean Atmosphere Data Set (COADS) for 1946-85. From Mitchell and
Wallace (1992).
power of harmonic analysis for detecting weak periodic signals.
2.2 Quasi-periodic phenomena
Because climate research tends to be driven, in large part, by the quest
for prediction, quasi-periodic phenomena have attracted a disproportionate
amount of attention per unit explained variance in the climate record. The
literature of the first half of this century abounds with claims of periodicities and quasi-periodicities in the climate record; many of them related
to the ll-year cycle in solar activity and its harmonics and subharmonics.
Few, if any of these claims have withstood the test oftime. Decidedly fewer
quasi-periodicities have been reported since the advent of power spectrum
analysis, which imposes more rigorous standards for assessing statistical
significance.
By far the most notable, and perhaps the only universally recognized
quasi-periodic phenomenon in the climate system on the interannual to
interdecadal time scale is the regular downward progression of alternating,
zonally symmetric easterly and westerly wind regimes in the equatorial