The World Ocean Database（WOD） is used to evaluate the halocline depth simulated by an ice-ocean coupled model in the Canada Basin during 1990–2008. Statistical results show that the simulated halocline is reliable....The World Ocean Database（WOD） is used to evaluate the halocline depth simulated by an ice-ocean coupled model in the Canada Basin during 1990–2008. Statistical results show that the simulated halocline is reliable.Comparing of the September sea ice extent between simulation and SSM/I dataset, a consistent interannual variability is found between them. Moreover, both the simulated and observed September sea ice extent show staircase declines in 2000–2008 compared to 1990–1999. That supports that the abrupt variations of the ocean surface stress curl anomaly in 2000–2008 are caused by rapid sea ice melting and also in favor of the realistic existence of the simulated variations. Responses to these changes can be found in the upper ocean circulation and the intermediate current variations in these two phases as well. The analysis shows that seasonal variations of the halocline are regulated by the seasonal variations of the Ekman pumping. On interannual time scale, the variations of the halocline have an inverse relationship with the ocean surface stress curl anomaly after 2000,while this relationship no longer applies in the 1990 s. It is pointed out that the regime shift in the Canada Basin can be derived to illustrate this phenomenon. Specifically, the halocline variations are dominated by advection in the 1990 s and Ekman pumping in the 2000 s respectively. Furthermore, the regime shift is caused by changing Transpolar Drift pathway and Ekman pumping area due to spatial deformation of the center Beaufort high（BH）relative to climatology.展开更多
基金The National Basic Research Program(973 Program)of China under contract No.2015CB953900the National Natural Science Foundation of China under contract No.41330960
文摘The World Ocean Database（WOD） is used to evaluate the halocline depth simulated by an ice-ocean coupled model in the Canada Basin during 1990–2008. Statistical results show that the simulated halocline is reliable.Comparing of the September sea ice extent between simulation and SSM/I dataset, a consistent interannual variability is found between them. Moreover, both the simulated and observed September sea ice extent show staircase declines in 2000–2008 compared to 1990–1999. That supports that the abrupt variations of the ocean surface stress curl anomaly in 2000–2008 are caused by rapid sea ice melting and also in favor of the realistic existence of the simulated variations. Responses to these changes can be found in the upper ocean circulation and the intermediate current variations in these two phases as well. The analysis shows that seasonal variations of the halocline are regulated by the seasonal variations of the Ekman pumping. On interannual time scale, the variations of the halocline have an inverse relationship with the ocean surface stress curl anomaly after 2000,while this relationship no longer applies in the 1990 s. It is pointed out that the regime shift in the Canada Basin can be derived to illustrate this phenomenon. Specifically, the halocline variations are dominated by advection in the 1990 s and Ekman pumping in the 2000 s respectively. Furthermore, the regime shift is caused by changing Transpolar Drift pathway and Ekman pumping area due to spatial deformation of the center Beaufort high（BH）relative to climatology.
