广东气象2024,Vol.46Issue(5) :15-20.DOI:10.3969/j.issn.1007-6190.2024.05.004

在模式参数有误差下使用四维变分同化的LAF法

The LAF Approach with 4D Variational Assimilation under Conditions of Model Parameter Errors

余晓健 庞绮汶
广东气象2024,Vol.46Issue(5) :15-20.DOI:10.3969/j.issn.1007-6190.2024.05.004

在模式参数有误差下使用四维变分同化的LAF法

The LAF Approach with 4D Variational Assimilation under Conditions of Model Parameter Errors

余晓健 1庞绮汶1
扫码查看

作者信息

  • 1. 佛山市南海区气象局,广东 佛山 528000
  • 折叠

摘要

为了有效利用各种观测资料改善观测误差来提高集合预报的效果,提出了一种新的预报方法,即拟在模式参数有误差的情况下,利用四维变分同化方法获取的初始场进行滞后时间集合预报(LAF).通过比较在不同的观测误差方差、不同观测个数以及不同观测间隔的条件下,新方法与传统的滞后时间集合预报方法和单考虑最优初始场的四维同化确定性预报的预报效果的差异.实验结果表明:使用四维变分同化分析场的滞后时间集合预报在短期内的预报效果较好,但随着预报时长的增加,预报效果与单滞后时间集合预报的预报效果相接近.

Abstract

To utilize efficiently various types of observational data to improve ensemble forecast,a new forecasting method was put forward in this paper in which lagged average ensemble forecasting(LAF)was made using the initial field acquired by 4DVar with the presence of model parameter errors(LAF-4DVar).Under the conditions of different observational error deviation,observational cases and observational inter-vals,differences were compared between deterministic forecast results which used either the traditional LAF,LAF-4DVar or 4DVar(taking into account the optimum initial field only).As shown in our experiments that the LAF-4DVar made the best forecast in the short term but was close to the traditional LAF with the increase of forecast validation time.

关键词

滞后时间集合预报(LAF)/四维变分同化/参数误差/最大简化气候模式

Key words

lagged average ensemble forecasting(LAF)/4-dimensional variational assimilation(4DVar)/parameter error/maximum simplified climate model

引用本文复制引用

基金项目

佛山市气象局2020年课题项目(202005)

出版年

2024
广东气象
广东省气象学会

广东气象

影响因子:1.013
ISSN:1007-6190
段落导航相关论文