Task 5 will rely on short-medium range temperature and precipitation probabilistic forecasts (4 weeks to 3 months) from the ECMWF, obtained by means of combining numerical weather prediction models used in IPMA (ECMWF-GFS) and statistical methods. The forecasts integrate near-real time observations and estimations. These forecasts will be used in order to produce new forecasted drought and soil moisture indicators (e.g., SPEI and SPI) which will be compared to observations. Moreover, short-medium range temperature and precipitation forecasts (5-10 days) also from the ECMWF will be used to develop to produce short-range forecasted drought and soil moisture indicators based on a moving average procedure.

On the other hand, obtaining climate change scenarios of climate extremes can be crucial for many sectors of the society, ranging from ecosystems and human activities to a range of economic sectors. In the context of this project, it is necessary to obtain reliable climate change projections on the likelihood evolution of HDE. It is expected that the new GCMs comparison exercise will be available in 2018 at time for IMPECAF, if not it will be necessary to use also global simulations from CMIP5 to evaluate the occurrence of HDEs for the European region. However, the intrinsic scale of the impacts associated with such extreme phenomena requires modelling at higher resolution scales, i.e. regional modelling.

The climate dataset to be used in this task corresponds to two high resolution WRF simulations (9km and 27km resolutions). These simulations will cover the entire IP and will be forced by EC-EARTH for present (1971-2000) and future (2071-2100) climate. Thus, this task has four main goals, namely is 1) to validate probabilistic short (5-10 days) and medium-range (4 weeks-3 months) operative meteorological forecasts from ECMWF, focused on the HDE identified in Task 2; 2) to draw climate change scenarios for the occurrence of HDE in the Iberia for the 21st century; 3) to assess how forecasts will affect yield losses in a short-medium range; and 4) to assess how predictions will affect yield losses in the future.​