IOD-related optimal initial errors and optimal precursors for IOD predictions from reanalysis data

详细信息    查看全文

• 作者：Rong Feng ; WanSuo Duan
• 关键词：Optimal precursors ; Initial errors ; Winter predictability barrier ; Indian Ocean dipole
• 刊名：Science China Earth Sciences
• 出版年：2017
• 出版时间：January 2017
• 年：2017
• 卷：60
• 期：1
• 页码：156-172
• 全文大小：
• 刊物主题：Earth Sciences, general;
• 出版者：Science China Press
• ISSN：1869-1897
• 卷排序：60

文摘

This study explored the spatial patterns of winter predictability barrier (WPB)-related optimal initial errors and optimal precursors for positive Indian Ocean dipole (IOD) events, and the associated physical mechanisms for their developments were analyzed using the Simple Ocean Data Assimilation dataset. Without consideration of the effects of model errors on “predictions,” it was assumed that different “predictions” are caused by different initial conditions. The two types of WPB-related optimal initial errors are almost opposite for the start months of July (–1) and July (0), although they both present a west-east dipole pattern in the tropical Indian Ocean, with the maximum errors located at the thermocline depth. Bjerknes feedback and ocean waves play important roles in the growth of prediction errors. These two physical mechanisms compete during July–December and ocean waves dominate during January–June. The spatial patterns of optimal precursors and the physical mechanisms for their developments are similar to those of WPB-related optimal initial errors. It is worth noting that large values of WPB-related optimal initial errors and optimal precursors are concentrated within a few locations, which probably represent the sensitive areas of targeted observations for positive IOD events. The great similarities between WPB-related optimal initial errors and optimal precursors suggest that were intensive observations performed over these areas, this would not only reduce initial errors and thus, prediction errors, but it would also permit the detection of the signal of IOD events in advance, greatly improving the forecast skill of positive IOD events.