The African easterly jet in the ECMWF Integrated Forecast System: 4D‐Var analysis

• Published in: Quarterly journal of the Royal Meteorological Society 2005-10, Vol.131 (611), p.2861-2885
• Main Authors: Tompkins, A. M., Diongue‐Niang, A., Parker, D. J., Thorncroft, C. D.
• Format: Article
• Language: English
• Subjects: 4D‐Var assimilation; Earth, ocean, space; Exact sciences and technology; External geophysics; JET2000 campaign; Meteorology; Weather analysis and prediction; Winds and their effects
• ISSN: 0035-9009; 1477-870X
• DOI: 10.1256/qj.04.136

In August 2000, the Met Office C‐130 research flight aircraft conducted four sorties over the western Sahel region, making dropsonde measurements at an unprecedented resolution. These were used as an independent validation source to assess the quality of the analysis of the African easterly jet (AEJ) in the ECMWF Integrated Forecast System (IFS). The model was found to produce a reasonably accurate representation of the AEJ location and structure. Peak zonal winds were within 5 m s−1 of the observed values for the two flights that performed transects across the jet. Temperature errors were less than 1 K above the boundary layer, and the gross features of the humidity distribution were reproduced. The analysis was also able to capture the weak African easterly wave activity present during the campaign, but was not capable of reproducing the observed daily variability in boundary‐layer height, with a too deep mixed layer on two days resulting in a significant dipole error structure in temperature and relative humidity. Analysis experiments were conducted in which various observations were excluded from the system to determine which were important for the diagnosis of the jet. Satellite data had a minor influence on the jet structure. Conventional platforms such as radiosondes and pilot balloons were more important. The thermodynamical information from radiosondes had a greater impact on the jet winds than the direct assimilation of the wind observations, made possible by the 4D variational assimilation system where thermodynamic information can impact on the dynamics. In contrast, the jet structure improved when surface observations (SYNOPs) were removed from the system, possibly due to the disparity between the radiosonde and SYNOP station humidity information over these regions. The investigation implied that all observational information is incorporated into the analysis in a well‐balanced manner, rather than being dominated by a single source. The system was able to assimilate the dropsonde data, which led to an improved dynamical and thermodynamical structure of the jet. Copyright © 2005 Royal Meteorological Society