JET2000 experiment : Synoptic analysis and NWP issues

Diongue, A., Parker, D.J., Thorncroft, C.D., Dumelow, R., Tompkins, A.M.

School of the Environment at the University of Leeds,UK

Abstract: The JET2000 experiment took place in the last week of August 2000 in West Africa. Four flights, involving transects along and across the African Easterly Jet (AEJ) and the baroclinic zone, were made using the UK Met Office C-130 Hercule aircraft. They deployed 111 dropsondes at 0.5-1 degree of resolution in addition to the aircfrat mesurements of the boundary layer. The AEJ, whose existence is due to the strong meridional surface temperature gradient over West Africa, is crucial both for generating synoptic-scale African easterly waves (AEWs) and organizing long-lived mesoscale convective systems (MCSs).

The JET2000 experiment provides a high resolution dataset allowing analysis of the meteorological features of the West African monsoon and their interactions on synoptic and diurnal time-scales. The flights observed an AEJ stronger and southward (20 m/s at 10 N) of climatological observations (15 m/s at 15N). Farther north in the Sahel, there is an evidence of a dry spell which may be consistent with a dry intrusion of mid-level air from the Sahara. The low-level observations of the meridional flights present significant mesoscale structures with overturning structures in the humidity and equivalent potential temperatutre fields. The boundary layer at the edge of the Sahara exhibit at late morning a split into a turbulent moist internal layer and a less turbulent dryer adaiabatic.

The ability of NWP models to analyse and forecast the features of the West African Monsoon region is also investigated, using the Unified Model of The Met Office and the ECMWF model. Preliminary results show that both models' analyses (without JET2000 extra data) appear to represent the AEJ well (in the second flight). However, the 5-day forecasts exhibit structures significantly different to the observations and climatology. To explore the significance of different data sources in generating a good analysis from a poor forecast over west Africa, sensitivity tests experiments are performed. In these experiments, the analysis without JET2000 extra data is considered as control and the routine observations are denied from a local window five days prior the date of the second flight.