<p>There is agreement on how the North Atlantic Oscillation forces the Atlantic Meridional Overturning Circulation, but the existence of a reversed interaction is widely disputed. Here, we investigate this type of ocean forcing the atmosphere by analysing several high- and low-resolution coupled climate models, ocean observations and reanalyses products of ocean and atmosphere. We find that in observations and about 50% of the coupled climate models, an ocean-forced negative North Atlantic Oscillation occurs at a lag of 5 years after the Atlantic Meridional Overturning Circulation peaks. Climate models with a strong cold temperature bias in the subpolar gyre and a positive sea-ice cover bias in the Atlantic and Arctic Ocean are unable to correctly simulate the heat flux pattern, resulting from the northward propagation of warm ocean temperatures, that forces the atmosphere. Efforts to remove this bias could therefore lead to substantial improvement in current decadal prediction systems.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Evidence for Atlantic Ocean forcing the atmosphere and the negative role of model bias

  • Rein Haarsma,
  • Sybren Drijfhout

摘要

There is agreement on how the North Atlantic Oscillation forces the Atlantic Meridional Overturning Circulation, but the existence of a reversed interaction is widely disputed. Here, we investigate this type of ocean forcing the atmosphere by analysing several high- and low-resolution coupled climate models, ocean observations and reanalyses products of ocean and atmosphere. We find that in observations and about 50% of the coupled climate models, an ocean-forced negative North Atlantic Oscillation occurs at a lag of 5 years after the Atlantic Meridional Overturning Circulation peaks. Climate models with a strong cold temperature bias in the subpolar gyre and a positive sea-ice cover bias in the Atlantic and Arctic Ocean are unable to correctly simulate the heat flux pattern, resulting from the northward propagation of warm ocean temperatures, that forces the atmosphere. Efforts to remove this bias could therefore lead to substantial improvement in current decadal prediction systems.