266
WILLIAM B. LARGE
variability even though it dominates the seasonal cycle. This notion
is confirmed by the annual anomalies of Q S . The maximum anomaly is
only −10W/m 2 . It is evident in 1992 between the equator and 20 ◦ S in
the Pacific and Atlantic Oceans, and to a lesser extent in the Indian.
This signal appears to be mostly a response to the Pinatubo volcanic
eruption of late 1991. The interannual variability of the air-sea freshwater flux is dominated by fluctuations in tropical precipitation. Therefore,
the interannual variability of freshwater forcing prior to the satellite era
remains obscure, because the dominant signal is missing.
The interannual AAO variability is most clearly reflected in the zonal
wind stress anomalies south of 40 ◦ S in all three basins, but especially
the Indian ocean. In particular, weaker eastward stress in the 1960s
corresponds to negative AAO, while the large positive AAO peaks in
1979, 1985, 1989, 1993 and 1998 are all associated with positive wind
stress anomalies. Meridional anomalies are weaker, but positive northward wind stress anomalies tend to be associated with negative AAO
(e.g. the 1960s). Interestingly, the changing winds are not consistently
reflected in the heat flux anomalies.
El Nino/Southern Oscillation (ENSO) variability is reflected in the
dominant tropical precipitation signals, with annual mean anomalies
from years following warm events being strongly positive in the Pacific,
weakly positive in the Indian, and negative in the Atlantic. The heat
fluxes have a corresponding signal only in the Pacific where they respond
directly to warmer SST by becoming more upward (negative). Thus,
they retard ENSO warming and cooling, as they do in a forced ocean
hindcast (e.g. Doney et al., 2003).
Zonal wind stress anomalies, and to a lesser extent meridional anomalies in the North Atlantic poleward of about 40 ◦ N reflect the North Atlantic Oscillation (NAO), with positive stress when the NAO index is
high. Corresponding negative heat flux anomalies (more latent and sensible cooling when the westerly winds are stronger) are evident, but less
pronounced during the period of generally high NAO index during the
1990s.
Neither the heat flux, nor wind stress anomalies in the North Pacific
appear to be directly related to the North Pacific Index (NPI). Low
values of the index correspond to both positive (e.g. 1970) and negative
(e.g. 1981 and 1983) zonal stress variations. The heat flux anomalies
tend to follow the wind and are, therefore, of the opposite sign. There
is no persistent anomaly in either the heat flux, or zonal stress through
the period of high NPI from 1989 through 1991.
WILLIAM B. LARGE
variability even though it dominates the seasonal cycle. This notion
is confirmed by the annual anomalies of Q S . The maximum anomaly is
only −10W/m 2 . It is evident in 1992 between the equator and 20 ◦ S in
the Pacific and Atlantic Oceans, and to a lesser extent in the Indian.
This signal appears to be mostly a response to the Pinatubo volcanic
eruption of late 1991. The interannual variability of the air-sea freshwater flux is dominated by fluctuations in tropical precipitation. Therefore,
the interannual variability of freshwater forcing prior to the satellite era
remains obscure, because the dominant signal is missing.
The interannual AAO variability is most clearly reflected in the zonal
wind stress anomalies south of 40 ◦ S in all three basins, but especially
the Indian ocean. In particular, weaker eastward stress in the 1960s
corresponds to negative AAO, while the large positive AAO peaks in
1979, 1985, 1989, 1993 and 1998 are all associated with positive wind
stress anomalies. Meridional anomalies are weaker, but positive northward wind stress anomalies tend to be associated with negative AAO
(e.g. the 1960s). Interestingly, the changing winds are not consistently
reflected in the heat flux anomalies.
El Nino/Southern Oscillation (ENSO) variability is reflected in the
dominant tropical precipitation signals, with annual mean anomalies
from years following warm events being strongly positive in the Pacific,
weakly positive in the Indian, and negative in the Atlantic. The heat
fluxes have a corresponding signal only in the Pacific where they respond
directly to warmer SST by becoming more upward (negative). Thus,
they retard ENSO warming and cooling, as they do in a forced ocean
hindcast (e.g. Doney et al., 2003).
Zonal wind stress anomalies, and to a lesser extent meridional anomalies in the North Atlantic poleward of about 40 ◦ N reflect the North Atlantic Oscillation (NAO), with positive stress when the NAO index is
high. Corresponding negative heat flux anomalies (more latent and sensible cooling when the westerly winds are stronger) are evident, but less
pronounced during the period of generally high NAO index during the
1990s.
Neither the heat flux, nor wind stress anomalies in the North Pacific
appear to be directly related to the North Pacific Index (NPI). Low
values of the index correspond to both positive (e.g. 1970) and negative
(e.g. 1981 and 1983) zonal stress variations. The heat flux anomalies
tend to follow the wind and are, therefore, of the opposite sign. There
is no persistent anomaly in either the heat flux, or zonal stress through
the period of high NPI from 1989 through 1991.