Premium
On the time variability of the net ocean‐to‐atmosphere heat flux in midlatitudes, with application to the North Atlantic basin
Author(s) -
Czaja Arnaud
Publication year - 2003
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1256/qj.02.66
Subject(s) - ocean gyre , climatology , forcing (mathematics) , ocean dynamics , ocean heat content , oceanic basin , mixed layer , atmosphere (unit) , middle latitudes , geostrophic wind , environmental science , flux (metallurgy) , ocean current , atmospheric sciences , geology , structural basin , meteorology , geography , materials science , fishery , metallurgy , biology , paleontology , subtropics
A new diagnostic to investigate the role of ocean dynamics in midlatitude air–sea interactions is presented and tested against observations. It is based on the analysis of the time variability of the net ocean‐to‐atmosphere heat flux ( F S ). A hierarchy of air–sea interaction models indicate that, in the absence of ocean dynamics, the power spectrum of F S should be blue at time‐scales longer than a threshold set by the late‐winter ocean mixed‐layer thickness and the sensitivity of F S to sea surface temperature anomalies. Comparison of the predicted F S spectrum with observations over the North Atlantic shows a good agreement over the subpolar gyre where the deep ocean mixed layer combined with strong stochastic forcing allows large fluctuations in F S at decadal and longer time‐scales. Discrepancies, however, arise over the Gulf Stream extension region. Here it is suggested that the observed variability of F S at time‐scales longer than a decade is controlled by geostrophic ocean dynamics rather than local atmospheric forcing. The diagnostic appears to be a useful and simple tool to investigate the role of ocean dynamics in the upper‐ocean heat budget. It is particularly well suited to the analysis of long simulations of coupled ocean–atmosphere models. One implication of the study for ocean‐only numerical simulations is that one cannot specify externally the low‐frequency variability of F S . The latter should only arise as a consequence of ocean dynamics. Copyright © 2003 Royal Meteorological Society