Impact of special features of numerically predicted and analysed meteorological data ont he results of ozone forecast by a pbl-ctm

摘要

The Freie Universitat Berlin and IVU~1develop a comple ozone forecast system, which combines statistical and fuzzy methods with regional Eulerian transport modelling. Each method is an independent forecast tool focussing on different spatial and temporal scales. The statistically and fuzzy based forecast is locally applied, whereas the Eulerian transport model predicts the large sale development of ozone patterns. The results of the transport modelling will be used as an additional parameter for the statistical and fuzzy methods in a framework of a neuro fuzzy moudule. Ideally, the combination of both approaches shall improve the forecast performance (Reimer and Dlabka, 1998).