14/07/11

Forecasters ‘warned of Horn of Africa drought’ last year

The drought is the worst in 60 years, according to the UN Copyright: Flickr/Oxfam International

Send to a friend

The details you provide on this page will not be used to send unsolicited email, and will not be sold to a 3rd party. See privacy policy.

[LONDON] Forecasting systems were warning about a serious drought in the Horn of Africa as much as a year ago — but communication problems between scientists and decision-makers meant the alerts went largely unheeded, according to forecasters.

Warnings about the drought — which the United Nations says is the worst in 60 years — were issued last August, when the Famine Early Warning Systems Network (FEWS NET) released a brief on food security in East Africa following the declaration of a La Niña event, a cooling of the sea surface in the Pacific Ocean known to affect weather in Africa.

"We were very confident that the October to December rains were going to be poor," Chris Hillbruner, a food security early warning specialist with FEWS NET, told SciDev.Net. "And there was an increased likelihood that the March to May rains were going to be poor as well."

Once the predictions for October–December proved correct, the agency started releasing food security alerts for the region in November, February, March, May and June, and organising multiagency meetings in Nairobi, in February, March and May.

The drought is now affecting ten million people in Djibouti, Ethiopia, Kenya, Somalia and Uganda.

Chris Funk, a climatologist with FEWS NET, said that the organisation’s experts have been "a little frustrated that we provided this information quite early" but not enough has been done to make good use of it.

"The technology has outpaced the response systems," he said. "We are still developing rapid response capability around that."

And Hillbruner added: "At the technical level, there’s been agreement about the situation for the past five months.

"What we’ve demonstrated over time, but particularly in the last year, is that systems have got to the point where they can do pretty good early warning within a timeframe that offers opportunities for response."

Randolph Kent, director of the Humanitarian Futures Program at the UK’s King’s College London, said that the drought response raised issues about how the humanitarian community turns scientific information and uncertain predictions into decisions about a response, particularly in turbulent regions.

His comments were supported by Molly Hellmuth, a researcher at the International Research Institute for Climate and Society at Columbia University in the United States, who co-edited a report about mitigating weather disasters, published last month in Climate and Society.

Hellmuth said clearer forecasts are needed that are aimed at decision-makers and come with recommended action steps stating what levels and types of action would be a sensible response.

And some ‘no regrets’ actions that may not have a negative effect even if the predictions do not come true could be encouraged — such as repositioning of food supplies and capacity building, she said.

Hillbruner agreed that the uncertainties in forecasting hamper its use: "In order to use early warning information most effectively, decision-makers have to be comfortable with that uncertainty — and it’s difficult to be comfortable with it."

But Funk added that the monitoring systems used for forecasting are now playing a powerful role in the huge humanitarian response now underway, making it a much more effective response than that to the Ethiopian famine in 1984–5.

Simon Mason, who works with Hellmuth, said that his research has shown that both forecasters and the end-users have a fear of being proved ‘wrong’ which causes all players to dilute their predictions and actions.

In a recent assessment of early warning systems, which did not include the FEWS NET Horn of Africa drought forecasts, Mason found that forecasters are "playing safe; being very reluctant to issue strong warnings of severe climate events even when their models are suggesting there is a high chance of more extreme conditions".

Hillbruner, however, denied that this had been the case with the Horn of Africa drought predictions.

In the humanitarian community, Andrew Collodel, an emergency programme coordinator with the aid agency HelpAge International agreed that the forecasts had been made and communicated. HelpAge had listened to early warnings and worked in the region to prepare for the drought, but it had failed to garner sufficient funds in advance of the crisis.

Donors had faced "confounding issues" such as the global financial situation, troubled local politics such as the chaotic situation in Somalia, and several other recent major crises that had raided their resources, such as the one following the Haiti earthquake and the tsunami in Japan.

"The early warning systems are not a problem, how we react to them is," he told SciDev.Net.

Link to FEWS NET’s initial August 2010 brief on food security in East Africa [219kB]

Link to full ‘A Better Climate for Disasters Risk Management’ [5.98MB]