The impact of April sea-ice thickness (SIT) initialization on the predictability of September sea-ice extent (SIE) is investigated based on a series of perfect model ensemble experiments using the MIROC5.2 climate model. Ensembles with April SIT initialization can accurately predict the September SIE for greater lead times than in cases without the initialization – up to 2 years ahead. The persistence of SIT correctly initialized in April contributes to the skilful prediction of SIE in the first September. On the other hand, errors in the initialization of SIT in April cause errors in the predicted sea-ice concentration and thickness in the Pacific sector from July to September and consequently influence the predictive skill with respect to SIE in September. The present study suggests that initialization of the April SIT in the Pacific sector significantly improves the accuracy of the September SIE forecasts by decreasing the errors in sea-ice fields from July to September.

中文翻译：

---

4 月初始化的海冰厚度对 MIROC 气候模型的北极海冰范围可预测性的影响

基于MIROC5.2气候模型的一系列完美模型集合实验，研究了4月海冰厚度（SIT）初始化对9月海冰范围（SIE）可预测性的影响。与没有初始化的情况相比，具有 4 月 SIT 初始化的集合可以准确预测 9 月的 SIE，提前 2 年。4 月正确初始化的 SIT 的持久性有助于对 9 月第一个 SIE 的熟练预测。另一方面，4 月 SIT 初始化的错误会导致 7 月至 9 月太平洋区域海冰浓度和厚度的预测误差，从而影响 9 月 SIE 的预测技巧。