The project Open Forecast (duration September 2018 to August 2020) is a novel demonstration project of the EU (CEF Telecom, Action Number 2017-DE-IA-0170). It will show how public and freely available data (open data) can be combined with supercomputer resources to create new data products for European citizens, public authorities, economic operators and decision makers. The data will be searchable and usable via services and the European Data Portal. Two use cases with specific processing chains are implemented within this project to demonstrate the applicability of the approach.
3D Particular Matter Forecast Calculations and Citizen Science Sensors
The following figure depicts the correlation between the immersive 3D climate model based forecasting simulation and the sensor data from the sensor.community network. For each sensor location, the time series of the actual measured data is compared to the computed values from the forecasting model at the same location.
High values in the northern part of the map depict a high correlation, meaning the temporal evaluation fits very nicely. Below the direct comparison of the measured and simulated data is shown.
The figures show that despite the known issues with the low cost sensors and the complexity of the 3D forecasting model, the measured data can actually be reproduced. A more detailed analysis will be published soon.
Open Forecast Services
To implement the different use cases following services are available as prototypes:
- Processing of open satellite data with the Sen2Agri-system. The service is available under https://sen2agri.open-forecast.eu/
- Providing the processed data via http://geoserver.gwdg.de/ , that is based on the open-source software Geoserver
- Design and run workflows using flowable under https://flow.open-forecast.eu
- API for sending jobs to the High Performance Computing infrastructure http://hpc-api.open-forecast.eu. This service is used with the workflow system. The API documentation is given here.