摘要
The characteristic low-frequency oscillation of the sea surface temperature anomaly (SSTA) of ENSO related regions, Nino 1 + 2, Nino 3, Nino 4 and Nino West, and the Southern Oscillation index (SOI) is analyzed with the method of maximum entropy spectrum. Antarctic sea ice is divided into 4 regions, i. e. East Antarctic is Region Ⅰ (0°-120° E), the region dominated by Ross Sea ice is Region Ⅱ (120° E-120° W), the region dominated by Ross Sea ice is Region Ⅲ (120° W-0°), and the whole Antarctic sea ice area is Region Ⅳ. Also, the month-to-month correlation series of the sea ice with ENSO from contemporary to 5-years lag is calculated. The optimum correlation period is selected from the series. The characteristics and the rules obtained are as follows.1. There are a common 4-years main period of the SSTA of Ninos 1 + 2,3 and 4, a rather strong 4-years secondary period and a quasi-8-years main period of that of Nino West. There are also 1. 5 and 2 to 3-years secondary periods of that of all 4 Nino regions.2. As another indicator of El Nino, the SOI represents the feature of the atmosphere in low latitude area, having a quasi-5-years main period; it also has secondary periods, 1, 1. 5 and 2 to 3 years, among them, the 2 to 3-years one is prominent.3. There is a close relationship between Antarctic sea ice and ENSO event. In the long-range correlation from contemporary up to 60 months of the SSTA in Ninos 1 + 2,3 and 4 and Antarctic sea ice area index, or the time series of 16 correlation coefficients made of each one of the 4 sea ice regions with that of the 4 Nino regions, there is a prominent common characteristic that all correlations are negative from contemporary to 34-months lag of the SSTA of the 4 Nino regions behind Antarctic sea ice, the optimum correlation periods with the confidences equal to or more than 5 % , 1 % lagging in 13-19, 24-34 month are the most. The correlations of sea ice in Regions Ⅱ , Ⅲ and W with the SSTA of Ninos 3 and 4 are the strongest. The correlation of the sea ice in Region Ⅰ with Nino West in 4 - 5-years lag becomes a very strong positive one. The correlations of the sea ice in Regions Ⅱ and Ⅲ with Ninos 1 + 2, 3 and 4 become strong positive ones during the quasi-4-years lag. The variation of the correlation series of the SOI and the 4 sea ice regions is the opposite of that of the 4 Nino regions. The correlations with the sea ice in Regions Ⅱ , Ⅲ and Ⅳ are strong, with the strong positive correlations of 6, 10 and 24-months lag being the optimum correlation periods. And the strong negative correlation period is 40-months lag.4. The characteristic variation of the correlation time series reflects the low-frequency oscillation feature of Antarctic sea ice and ENSO. In the periodic variation, the correlation becomes the strongest when the ice and ENSO are inresonance. Specifically,the Antarctic sea ice influences ENSO most in an earlier period of its own variation. Moreover, it is also related with the period of variation of ENSO itself, i. e. the correlation of sea ice and ENSO gets the best in a period lag of ENSO its own variation.
The characteristic low-frequency oscillation of the sea surface temperature anomaly (SSTA) of ENSO related regions, Nino 1 + 2, Nino 3, Nino 4 and Nino West, and the Southern Oscillation index (SOI) is analyzed with the method of maximum entropy spectrum. Antarctic sea ice is divided into 4 regions, i. e. East Antarctic is Region Ⅰ (0°-120° E), the region dominated by Ross Sea ice is Region Ⅱ (120° E-120° W), the region dominated by Ross Sea ice is Region Ⅲ (120° W-0°), and the whole Antarctic sea ice area is Region Ⅳ. Also, the month-to-month correlation series of the sea ice with ENSO from contemporary to 5-years lag is calculated. The optimum correlation period is selected from the series. The characteristics and the rules obtained are as follows.1. There are a common 4-years main period of the SSTA of Ninos 1 + 2,3 and 4, a rather strong 4-years secondary period and a quasi-8-years main period of that of Nino West. There are also 1. 5 and 2 to 3-years secondary periods of that of all 4 Nino regions.2. As another indicator of El Nino, the SOI represents the feature of the atmosphere in low latitude area, having a quasi-5-years main period; it also has secondary periods, 1, 1. 5 and 2 to 3 years, among them, the 2 to 3-years one is prominent.3. There is a close relationship between Antarctic sea ice and ENSO event. In the long-range correlation from contemporary up to 60 months of the SSTA in Ninos 1 + 2,3 and 4 and Antarctic sea ice area index, or the time series of 16 correlation coefficients made of each one of the 4 sea ice regions with that of the 4 Nino regions, there is a prominent common characteristic that all correlations are negative from contemporary to 34-months lag of the SSTA of the 4 Nino regions behind Antarctic sea ice, the optimum correlation periods with the confidences equal to or more than 5 % , 1 % lagging in 13-19, 24-34 month are the most. The correlations of sea ice in Regions Ⅱ , Ⅲ and W with the SSTA of Ninos 3 and 4 are the strongest. The correlation of the sea ice in Region Ⅰ with Nino West in 4 - 5-years lag becomes a very strong positive one. The correlations of the sea ice in Regions Ⅱ and Ⅲ with Ninos 1 + 2, 3 and 4 become strong positive ones during the quasi-4-years lag. The variation of the correlation series of the SOI and the 4 sea ice regions is the opposite of that of the 4 Nino regions. The correlations with the sea ice in Regions Ⅱ , Ⅲ and Ⅳ are strong, with the strong positive correlations of 6, 10 and 24-months lag being the optimum correlation periods. And the strong negative correlation period is 40-months lag.4. The characteristic variation of the correlation time series reflects the low-frequency oscillation feature of Antarctic sea ice and ENSO. In the periodic variation, the correlation becomes the strongest when the ice and ENSO are inresonance. Specifically,the Antarctic sea ice influences ENSO most in an earlier period of its own variation. Moreover, it is also related with the period of variation of ENSO itself, i. e. the correlation of sea ice and ENSO gets the best in a period lag of ENSO its own variation.
