Creating More Useful Forecasts

Seasonal forecasts can be effective tools for agricultural planners, water resources managers and other decision makers. For example, after torrential rains and floods wreaked havoc in the West African nation of Ghana in 2007, displacing some 400,000 people there, the regional office of the International Federation of Red Cross and Red Crescent Societies started using seasonal rainfall forecasts as part of its planning operations. When similar flooding occurred in 2008, the office was well-prepared and was able to save lives and livelihoods, as well as reduce the costs of providing relief to those affected.

Despite major advances in seasonal forecasting capabilities, the use of these forecasts is still fairly limited, especially in developing countries, where climate scientists often lack the resources or expertise to customize forecasts for specific needs. To address this issue, the IRI and its partners have led training workshops for professionals around the world for the last eight years. One recently took place in Beijing, China.

Twenty five experts from meteorological and climate agencies in 17 countries took part in the two-week workshop, which was sponsored by the IRI and the World Meteorological Organization and hosted by the Beijing Climate Center. The participants learned how to use statistical approaches to seasonal forecasting and how to "downscale"-- which is the process of teasing out regional and local details from the coarse resolution of global climate models. Just as importantly, the participants learned how to more effectively communicate forecasts and forecast quality to users such as health and agriculture planners back home.

"This workshop was our most important one to date," says Simon Mason, who runs IRI's Climate Program. "With the support of the WMO, we've aimed to develop experts around the world who could lead or assist in subsequent training activities, and who could provide technical support to colleagues in their region."

Central to the training was the Climate Predictability Tool (CPT), IRI's free downloadable application for making tailored seasonal climate forecasts (see earlier story). Mason and Michael Tippett created the CPT in 2002 specifically for meteorologists and other forecasters in developing countries who don't have access to powerful computers and other resources. The tool allows them to make usable, customized forecasts, complete with maps and charts, in a matter of minutes. It has been downloaded more than 1,200 times last year alone and has a strong global user community.

Mason and the other trainers encouraged the forecasters to think about how they could present forecasts in a way that was most beneficial to their users. "National meteorological services in many countries could greatly improve their effectiveness by focusing more directly on the questions users ask of them rather than expecting the users to understand the default forecast formats," he says.

The participants in Beijing were treated to a new version of the CPT that included many enhancements requested by users around the world. In fact, Mason says most of the changes made to the program since its creation have come from the user community. "The CPT is designed for operational forecasters, who often have to make regular predictions with limited resources and minimal time. We pay close attention to their suggestions so that the software provides them with the outputs they need," he says.

Close attention indeed: during the workshop, one of the participants, Philip Aming'o Omondi from Kenya, thought it would be useful if he could compare forecasts to conditions that occurred in definable 'base' years. Within a couple of days, Mason had rolled this feature into the software. He also tweaked the program so that it was more compatible with Spanish and Portuguese versions of the Windows operating system. Thanks to Omondi's suggestion and Mason's quick response, another participant, Juan Jose Nieto, was able to generate a forecast for his native Ecuador that showed how conditions in the upcoming rainy season compared to those experienced in 2007 and 2008. This feature was so useful that Nieto, who works at International Research Centre on El Niño (CIIFEN), was able to convince national meteorological services in the region to begin issuing some of their forecasts in this manner-something which was easy to implement because many of them are already using CPT to produce their operational seasonal forecasts.

The recent IRI-WMO Workshop participants now join a set of worldwide trainers and regional experts in producing high quality, tailored forecasts in user-friendly formats. Like Nieto and Omondi, they're passing this knowledge on to their colleagues. These are vital steps in strengthening the communication of climate information to climate-sensitive sectors such as public health, and agriculture, ultimately enabling more effective climate-related risk management practices.