Scholarlink

<p>Recently, a coupled data assimilation system based on the community earth system model (CESM) and ensemble adjustment Kalman filter (EAKF) has been established to assimilate various ocean observations including gridded sea surface temperature and in situ temperature and salinity profiles for the initialization of seasonal prediction. The main goal of the present study is to assess the El Niño-Southern Oscillation (ENSO) prediction capability of the newly developed system (CESM-E). We compare it with a benchmark prediction system based on the same model but employing a nudging scheme (CESM-N), which nudged the wind fields and ocean temperature. Results have found that although the initial subsurface temperature are comparable in the two systems, CESM-E outperforms CESM-N in a few aspects. For example, CESM-E exhibits clearly lower root mean square errors in the first few leading months and higher anomaly correlation coefficients in the Niño4 region. In addition, case studies reveal that CESM-E is clearly better in predicting the 2006/2007 El Niño and 2010/2011 La Niña events. Reasons behind the improvement of CESM-E are studied, which can provide useful insights into the design of a data assimilation system and the further improvement of current ENSO prediction system.</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

Assessment of an ENSO prediction system based on the community earth system model and ensemble adjustment Kalman filter

  • 期刊论文

  • 发表日期：

    2025年9月1日

查看全文
  • Yihao Chen,
  • Zheqi Shen,
  • Xunshu Song

摘要

Recently, a coupled data assimilation system based on the community earth system model (CESM) and ensemble adjustment Kalman filter (EAKF) has been established to assimilate various ocean observations including gridded sea surface temperature and in situ temperature and salinity profiles for the initialization of seasonal prediction. The main goal of the present study is to assess the El Niño-Southern Oscillation (ENSO) prediction capability of the newly developed system (CESM-E). We compare it with a benchmark prediction system based on the same model but employing a nudging scheme (CESM-N), which nudged the wind fields and ocean temperature. Results have found that although the initial subsurface temperature are comparable in the two systems, CESM-E outperforms CESM-N in a few aspects. For example, CESM-E exhibits clearly lower root mean square errors in the first few leading months and higher anomaly correlation coefficients in the Niño4 region. In addition, case studies reveal that CESM-E is clearly better in predicting the 2006/2007 El Niño and 2010/2011 La Niña events. Reasons behind the improvement of CESM-E are studied, which can provide useful insights into the design of a data assimilation system and the further improvement of current ENSO prediction system.