Air Quality Forecasting System: Needs of Computational Resources and Database
Jose M. Baldasano
Earth Science Department, Barcelona Supercomputing Center-Centro Nacional de Supercomputacion (BSC-CNS)
Environmental Modeling Laboratory, Technical University of Catalonia (UPC)
Barcelona, Spain
Abstract
The systems of air quality analysis historically have moved from monitoring networks to be supplemented by air quality models. In the last ten years, have focused very quickly towards air quality forecasting systems (AQFS), with a high degree of complexity in modeling aspects and the amount of data they use; emissions, resolutions from large domains to the local scale, requiring progressively greater computational resources. In addition, working with input data and results in real-time, which requires signigicant communication and web resources; as well as the need to use powerful visualization systems.
The presentation would be based mainly on the experience with the "Spanish Air Quality Forecasting System CALIOPE" (www.bsc.es/caliope) developed and operated in the department of Earth Sciences of BSC, supplemented by information from other AQFS operating in Europe (France, Germany, UK, Belgium, etc.). The presentation will focus on four main points: 1) Computational issue, 2) Data Management, 3) Communications and data transfer for data assimilation, and 4) Visualization.