Impact of tropospheric and stratospheric data assimilation on mesospheric prediction

Abstract

Numerical experiments are used to assess the potential benefit of the assimilation of tropospheric and stratospheric observations on mesospheric prediction. A simulated atmosphere, taken as truth, is created using the Canadian Middle Atmosphere Model (CMAM). The truth is sampled at the locations of the measurements from the actual observing system to produce observations, which are then assimilated with theCMAM-DAS (Data Assimilation System). Obtained forecasts are compared with the truth, and error statistics are calculated. An assessment based on predictability shows that upward propagation of information resulting from the assimilation of tropospheric and stratospheric observations improves the mesosphere in the largest scales (with horizontal wavenumbers less than approximately 10). At the same time, the principal inability of the system to predict mesospheric small scales is demonstrated.Numerical experiments are used to assess the potential benefit of the assimilation of tropospheric and stratospheric observations on mesospheric prediction. A simulated atmosphere, taken as truth, is created using the Canadian Middle Atmosphere Model (CMAM). The truth is sampled at the locations of the measurements from the actual observing system to produce observations, which are then assimilated with theCMAM-DAS (Data Assimilation System). Obtained forecasts are compared with the truth, and error statistics are calculated. An assessment based on predictability shows that upward propagation of information resulting from the assimilation of tropospheric and stratospheric observations improves the mesosphere in the largest scales (with horizontal wavenumbers less than approximately 10). At the same time, the principal inability of the system to predict mesospheric small scales is demonstrated.