How good are the ECMWF models at forecasting the probability of warmer than average conditions, such as the heatwave that gripped Europe this summer?
Extended range forecasts have been a difficult product to put much faith in, as a forecast extends out in time there is a tendency towards chaos, however in recent years with growing computing power and understanding the ability to gather useful information from medium range forecasts has increased.
The ability to gain confidence in a forecast like this would allow the farming community, among others, to prepare in advance for such conditions. Economic stresses could be reduced, water, energy and disaster plans could be prepared in advance.
ECMWF’s forecast at three weeks out, and to a slighter extend four weeks out, showed a positive temperature bias at the beginning of August suggesting a prolonged period of above average temperatures. While there were some notable errors over Spain, value can be gained from the trends in this information.
The following article goes into detail on the global ingredients, such as the Madden Julian Oscillation(MJO), and high atmosphere warming events monitored to produce a medium range forecast. An insight is given into the process of delivering the highest skill model data to deliver the medium range forecast, as well as current and future uses of the extended range forecasts.