A feasible approach to improve forecast skill of summer precipitation in northeast China by statistical regression of the northeast China cold vortex in the multimodel ensemble. (27th May 2021)

• Record Type: Journal Article
• Authors: Ding, Ting; Gao, Hui
• Is Part Of: International journal of climatology. Volume 41:Number 14(2021)
• Page Start: 6397
• Page End: 6414
• Publication Date: 2021-05-27
• Subjects: equally weighted ensemble mean -- linear regression -- multimodel -- northeast China cold vortex -- precipitation; Climatology -- Periodicals; Climat -- Périodiques; Climatologie -- Périodiques; 551.605
• DOI: 10.1002/joc.7202
• Languages: English
• ISSNs: 0899-8418

Abstract

Abstract: Hindcasts during 1991–2010 are analysed to investigate the forecast skills for summer precipitation in northeast China (NEC) and circulations in northeast Asia with a leading time of 1‐month in three operational climate models and their equally weighted mean. It is found that the precipitation prediction ability of all the models and their combinations are much low in NEC. However, both the single model and their averages can well reproduce the basic summer circulation characteristics over Northeast Asia, especially the northeast China cold vortex (NECV) that directly affects precipitation in NEC. Observed precipitation is significantly correlated with the forecasted vorticities in each model and their combinations, particularly in the ensemble mean of the models (MME). Furthermore, the interannual variability of the NECV in MME agrees well with the observations. Linear regression between the simulated NECV and the precipitation in NEC is established for each model and their averages. The root‐mean‐square errors (RMSEs) of precipitation forecast from the regression of NECV are much lower than the direct outputs in above models, while the correlation coefficients (CCs) are much higher. The CC between observation and forecasts by the regression method based on the NECV in MME is the highest (0.5), closest to the value of 0.58 in observation, and the RMSE is the smallest (0.54), decreased by 13% compared with MME direct outputs. The cross‐validation during the 20 years can provide better prediction skills with a CC of 0.37, higher than that for the single‐model outputs, the averages of random two models and the MME. Independent forecasts also provide higher prediction skills than the official forecast of China Meteorological Administration in 4 years during 2013–2018, especially the extreme flood summer of 2013. The results indicate it might be a feasible approach to improve the prediction skill of summer precipitation in NEC. Abstract : The precipitation prediction abilities of three operational climate models and their combinations are much low in northeast China (NEC). A linear regression forecast between the simulated northeast China cold vortex with a leading time of 1 month and the precipitation in NEC is established for each individual model and multimodel ensemble. This can improve the prediction skill of summer precipitation in NEC.