During the last years the frequency of extreme weather events are increasing due to Climate Change and Global Warming. The Climate variations are affecting to global population in such a way that the discussion is in the worldwide agenda. The agri-food sector is one of the most affected due to climate variations by the extreme events such as hail or frost. Important advances in the global climate models have occurred along the last years thanks to technical improvements in the computational sector. However, some of these good results cannot be used by users beyond the scientific or decision-making community due to complexity to reach and interpretation the Climate Information. Thanks to R+D projects funded by governments or supranational communities, part of the advances and new technologies can be exploited by a large number of user thanks to collaborations among the technology industries, academic community, commercial sector, potential end-users and other relevant actors.
GMV is participating in CYBELE, an EU-funded project that aims to generate innovation and create value in the domain of agri-food by applying Precision Agriculture (PA) and Precision Livestock Farming (PLF) methods.
Right from the project kickoff GMV has been leading one of the nine demonstrators to weigh up and show the usefulness of PA technology, concentrating on the development of climate services for organic fruit production. This prototype is fruit of the first 2 years of project development, combining HPC hardware and software and Big-Data for running scalable data analyses.
GMV is concern about the data-processing challenges involved in PA, plus solutions and optimizations in the implementation of field-scale yield data analysis using the CYBELE platform.
Working on this demonstrator, GMV has achieved notable headway in the conceptualization and exploration of machine-learning methods for forecasting frost and hail. These demonstrator trials were carried out in two zones of Valencia (Spain) region.
This Spanish region have suffered the negative impact by hail and frost events. According to Agroseguro (Spanish agricultural insurance system) paid 14.3 million euros in compensation for claims produced in persimmon crops in some Valencian municipalities due to damage by hail in November 2020.
Due to the sheer complexity of forecasting extreme-weather events, the trials were broken down into three blocks: hail forecasting, frost forecasting and the phenology of each type of fruit.
Each development implied efforts depending on particular cases however; a further experimentation process and a continuous communication with CYBELE’ technical partners (CINECA) and the stakeholder (CACV) have allowed to reach some encouraging results to think in an useful tool to mitigate the damages by extreme weather events by using HPC technology and Machine Learning.
GMV’s work on this demonstrator draws on data from the Meteorological Archival and Retrieval System of the European Centre for Medium-Range Weather Forecasts (ECMWF) and the Spanish Meteorological Agency (Agencia Española de Meteorología: AEMET).