335
CO2 concentration. If the CO2 content is gradually quadrupled, the Atlantic thermohaline circulation ceases and never recovers.
In the Manabe and Stouffer (1994) experiments, the weakening of the
thermohaline circulation is caused by the increase in atmospheric temperature and hence water vapor content, leading to enhanced poleward moisture
transport and lower high-latitude salinities. But the resilience of NADW
formation to the increased freshwater input in the sinking region depends
on various feedback processes, which are possibly incorrectly represented
in the coupled general circulation model (GCM), as demonstrated by the
large flux adjustments applied to model current climate. There is a need
to identify the important processes and assess how well any comprehensive
model represents them.
Here, I use the simplest possible model of interactions between atmospheric transports and the thermohaline circulation to analyze some feedback processes that I believe are important. Different model versions employ different assumptions about atmospheric transports, which are shown
to correspond to different stages of ocean-atmosphere coupling. I then
investigate the transient behavior of the coupled models near the steady
state corresponding to vigorous NADW formation.
The model builds upon the ones developed earlier by Nakamura, Stone,
and Marotzke (1994, NSM hereinafter) and by Marotzke and Stone (1995,
MS hereinafter), but differs in that a greater variety of atmospheric transport formulations is analyzed in detail. The paper is organised as follows.
Section 2 presents the model, closely following MS. Section 3 discusses the
stability of the equilibrium corresponding to vigorous N ADW formation, by
performing numerical integrations of the various model versions. Section
4 discusses the effects that land can have on ocean-atmosphere coupling,
and section 5 addresses climate drifts as seen in coupled GCMs despite the
application of flux adjustments. Section 6 analyzes a particular type of
variability, called 'neutrally buoyant mode' by Saravanan and McWilliams
(1995), in terms of the superposition of non-orthogonal eigenfunctions. A
few concluding remarks follow in section 7. Section 3 revisits topics discussed earlier in NSM and MS, but the discussion of competing feedbacks
is more definitive. The material in sections 4, 5, and 6 is novel.
CO2 concentration. If the CO2 content is gradually quadrupled, the Atlantic thermohaline circulation ceases and never recovers.
In the Manabe and Stouffer (1994) experiments, the weakening of the
thermohaline circulation is caused by the increase in atmospheric temperature and hence water vapor content, leading to enhanced poleward moisture
transport and lower high-latitude salinities. But the resilience of NADW
formation to the increased freshwater input in the sinking region depends
on various feedback processes, which are possibly incorrectly represented
in the coupled general circulation model (GCM), as demonstrated by the
large flux adjustments applied to model current climate. There is a need
to identify the important processes and assess how well any comprehensive
model represents them.
Here, I use the simplest possible model of interactions between atmospheric transports and the thermohaline circulation to analyze some feedback processes that I believe are important. Different model versions employ different assumptions about atmospheric transports, which are shown
to correspond to different stages of ocean-atmosphere coupling. I then
investigate the transient behavior of the coupled models near the steady
state corresponding to vigorous NADW formation.
The model builds upon the ones developed earlier by Nakamura, Stone,
and Marotzke (1994, NSM hereinafter) and by Marotzke and Stone (1995,
MS hereinafter), but differs in that a greater variety of atmospheric transport formulations is analyzed in detail. The paper is organised as follows.
Section 2 presents the model, closely following MS. Section 3 discusses the
stability of the equilibrium corresponding to vigorous N ADW formation, by
performing numerical integrations of the various model versions. Section
4 discusses the effects that land can have on ocean-atmosphere coupling,
and section 5 addresses climate drifts as seen in coupled GCMs despite the
application of flux adjustments. Section 6 analyzes a particular type of
variability, called 'neutrally buoyant mode' by Saravanan and McWilliams
(1995), in terms of the superposition of non-orthogonal eigenfunctions. A
few concluding remarks follow in section 7. Section 3 revisits topics discussed earlier in NSM and MS, but the discussion of competing feedbacks
is more definitive. The material in sections 4, 5, and 6 is novel.