Environmental modeling in an event-driven multitasking network environment

In order to provide fast analysis and forecast of a future state of a considered system, environmentalmodeling requires integration of different tools, data and methods. Using the example of the semi-empiricalbiogenic volatile organic compound emission model (seBVOC) the integration of numerical weather predictionmodel (NWP) output, Geographical Information System (GIS) and the seBVOC model with an event-drivenmultitasking network program is demonstrated. It is based on the Perl Object Environment (POE) and includesHTTP data request and download,Web services exposing GIS functionalities and the model run. The system isrun in parallel and in a collaborative event-driven way on several computer platforms connected to a network.Compared to the sequential approach a speed-up factor of up to 2 has been reached in the seBVOC applicationin the NATAIR project.