Prediction of African waves and specification of squall lines

This paper is part of a sequence of numerical weather prediction efforts utilizing the special observations from the GARP Atlantic Tropical Experiment (GATE). These observations, along with those available from the World Weather Watch, provide a unique opportunity to carry out numerical weather prediction (NWP) studies over portions of West Africa and the eastern Atlantic ocean utilizing high resolution limited area models. The present study reports on the results of 34 short-range numerical weather prediction experiments, with two simple models, designed to study the westward propagation of African waves. In the time frame of 24 to 48 hours, predictions at 700 mb (the level of maximum amplitude of these waves, and the level of non-divergence) show extremely promising results. These results are comparable to fine-mesh predictions in higher latitudes. Errors in phase speed and vector root mean square (RMS) wind errors are summarized for the third phase of GATE. A second part of this study deals with the problem of specification of the West African squall line with respect to the position of the predicted trough line of the African wave. A body of recent literature provides useful observational statistics relating the position of African waves, certain geographical regions, the time of day and the formation of line squall systems. These studies also provide information on the relative speeds of westward motion of the squall line and the wave trough. The decay of the squall lines during their westward propagation away from the trough is also a part of these observational statistics. Based on these observations, an empirical set of selection rules may be defined to specify the squall lines with respect to the predicted motion field. Successful results were obtained in the time frame of 24 to 48 hours for these experiments, the results of which are summarized below.