Medium Range Heat Health Forecasting

Heat health warning systems with a forecast horizon of up to two days have been established in several countries during the last decade. In addition, it seems useful to complement a heat-health warning system by an advance heat information that contains a corresponding forecast with lead times of up to 10 days. Unfortunately, the uncertainty of a forecast increases with its lead time. Probabilistic forecasts can be used to estimate this uncertainty of medium range weather forecasting.

Therefore the aim of this study was to develop a system which is able to give a hint at the possibility of a heat wave occuring within the next ten days. Probability forecasts were computed based on the ECMWF ensemble prediction system (EPS) with 50 ensemble members and one control run for each model run.

Three strategies are being tested. The first is to use only the 850-hPa air temperature. The second is to use the 2 m air temperature and the third to combine the ensemble forecasts with a complete heat budget model of a human being which takes into account all heat exchange mechanisms. The input variables for the heat budget model include air temperature, dew point, wind speed, and cloud coverage in order to calculate the radiant fluxes. The skills of these forecasts were tested using the Brier Score and the Brier Skill Score. In order to identify situations with a potential danger for human health, adaptation to the local weather situation of the last months was taken into account.

In a first step the mentioned strategies are used to simulate the 2003 heat wave in western Europe.