Summary
The skill in weather forecasting has been improving continuously for more than three decades due to advances in the scientific understanding of climate drivers, in the utilization of improved and extended observations, and in supercomputing capacity. Predictive skill has been increasing at a rate of one day per decade for short-range forecasts, so that current 6-day forecast is as accurate as the 5-day forecast ten years ago. Substantial further progress in reliability and skill in weather forecasting is expected in the future. Seasonal forecasts that provide predictions of how the average atmospheric, ocean and land surface conditions over particular areas and periods of time are likely to be different from the long-term average undergo a comparable development. Improving the skill of seasonal forecasts over Europe is always challenging because average predictability is low. Skill comes from capturing both interannual variations and long-term trends, and while it is often low over Europe on average, particularly in winter, reliability is generally quite high in the most recent ECMWF (European Centre for Medium-Range Weather Forecasts) long-range forecasting system SEAS5.
Despite increasing appreciation of the potential value of weather forecasts for agriculture, there is still a gap between what scientists consider as “useful” information and what users (e.g. farmers, advisory services, policy makers) recognise as “usable” in their decision-making processes. This gap is largely due to (i) difficulties in the initialisation of numerical weather prediction models and the spatial downscaling of forecasts to match farm- and field-level needs due to, among others, lack of spatially-dense network of weather stations; (ii) inadequate tailoring (i.e. translating, formatting, and communicating) of weather forecasts into more useful and usable forms for supporting context-specific decisions especially with regard to probabilistic forecasting systems; and (iii) insufficient level of interaction between providers and users of forecasts in order to co-develop and evaluate tailored information and, thereby, improve users perceptions of information salience, credibility, and legitimacy in particular decision contexts.
To overcome gap (i), it is planned to use both a statistical and a dynamical downscaling method to bring the information from the global model grid (36x36km grid box size) down to the 1x1km INCA (Integrated Nowcasting through Comprehensive Analysis) grid to generate spatially homogeneous ensemble weather forecast data sets. This approach with two different downscaling methods is chosen as both approaches have been extensively used and critically assessed in climate change studies, but still their added value for seasonal forecasting is not well understood yet. The INCA system, developed by ZAMG, is based on blending observations and numerical prediction (NWP) model fields. It also exploits remote sensing information (such as radar and satellite data) as well as high resolution time invariant information like topography and surface type. The INCA system provides frequently updated analyses and forecasts in the nowcasting range. A new approach that will be applied in the project is the combination of INCA data (from beginning of the growing season up to now) and downscaled probabilistic forecast data (up to +7 months in the future) which will lead to a spatial consistent data set as input for the crop models.
To address the inadequate tailoring of weather forecasts into more “usable” forms, the basic probabilistic forecast data need to be translated into tailored forecast products for specific set of decisions in order to increase efficient use of resources (e.g. timing and amount of crop specific fertilization and irrigation). Crop simulation models provide a means for translating forecasts into new and more relevant information for stakeholders involved in the agricultural sector. For instance, accurate prediction of crop phenological development by linking weather forecasts with crop models is valuable to farmers because many in-season management operations are commonly scheduled according to crop phenology.