Weather wise

A CONFERENCE is taking place this week in Abidjan in the Ivory Coast called "The First Forum on Seasonal Climate Prediction and its Application to Early Warning Systems for Food Security in West Africa". Its main aim is to coordinate forecasting techniques and assemble and analyse available data to produce a series of models to predict crop yields in the region.

A Reuters report on the first day of the conference indicates the potential that already exists to improve matters, but also identifies factors that may frustrate the best-intentioned efforts. On the positive side, delegates have reported how pilot projects in Asia with groundnut farmers could serve as a model for West Africa's cassava growers and may then be expanded to other crops.

Weather models may also help produce more accurate regional forecasts, which can be adapted to farmers' needs. Much historical weather data also exists that has never been collated and analysed. Research has already shown, for example, a link between drought in Somalia and high surface temperatures in the Atlantic.

All this offers hope, but, the big problem is El Nino. In fact, it is two problems in one. First, nobody can confidently predict when the next El Nino will be along or how big it will be; second, nobody can predict its precise effects. All we know is that once every five or so years, something may happen to wreck a nation's entire crop. Our understanding of weather needs to move forward a long way before the ambitions of this conference are fully realised.