Understanding African Easterly Waves and their Predictability: a moist singular vector appraoch

The easterly waves that occur in West African summer monsoon are of vital importance to the region as the convective rainfall is mostly organised by them. The variability of this rainfall can have devastating impacts on the already vulnerable population. A small number of these weather systems transform downstream in the north Atlantic into hurricanes. However, despite the significance of these easterly waves, weather and climate models continue to have difficulties in predicting them and simulating the seasonal rainfall distribution in the region. A technique has been developed to find small changes to initial data, 'optimal perturbations', that produce the largest changes in the weather predicted in, for example, one day. This technique has been used to help understand the growth of high impact storms in middle latitudes. It is also the basis for a scheme that provides a range of slightly different initial conditions for a forecast system that produces an ensemble of weather forecasts. This ensemble gives indications of the probability of future weather events. Optimal perturbations are also used for giving indications of the sensitivity of tropical cyclone behaviour, but the technique has otherwise not been used for the tropics. In this project, a new version of the package to produce these optimal perturbations will be applied to the easterly waves in West Africa. It will be used to help the theoretical understanding of these systems. This will give a basis for using the important new observations that will be taken in 2006 in a series of international campaigns. It will also show how predictable African easterly waves are, including the convective rainfall they give and the likelihood that they will turn in to a hurricane. The project will provide a basis for weather forecast centres to construct an ensemble forecast system that is focussed initially on the tropics, but may also help in suggesting subsequent changes in middle latitudes.