For some decades now, summer smog has been recognized as a severe environmental problem, and intensive research has been done to understand this phenomenon and to give advice to officials (Finlayson-Pitts and Pitts; 1986). Very soon, it became clear that complex interactions of meteorological, chemical, biological, and anthropogenic processes influence the formation of chemical photooxidants. To judge their relative importance for the creation and destruction potential of hazardous species like ozone, and to derive quantitative statements, computer models were found to be very important. Many numerical models, therefore have been implemented, and numerical techniques have evolved rapidly.
First modeling attempts focused mainly on photochemical reactions, or they concentrated on air dispersion modeling and omitted the chemical transformation, but an increasing number of models tried to consider the coupling of meteorological and chemical processes. To account for horizontal transport, vertical mixing, and to assess the spatial distribution of ozone, three-dimensional Eulerian models such as CIT (McRae et al.; 1982, 1983) or UAM (Environmental Protection Agency; 1980; Scheffe and Morris; 1993) - so called urban airshed models - were developed. They require values for wind and turbulence as drivers for the model. Many studies were carried out with spatially interpolated data measured at a few fixed locations. These procedures, however, are limited in their ability to represent dynamically changing situations which occur especially in complex terrain. As an alternative, mesoscale meteorological models can be used to produce wind and turbulence fields, (see e.g: TVM , Thunis; 1995; METRAS , Schluenzen et al.; 1996; KAMM , Vogel et al.; 1987; Adrea , Bartzis et al.; 1993; MM5 , Grell et al.; 1993; Mercure , Buty; 1988). Unfortunately, the decoupled application of a meteorological and a chemical program risks to be error prone because of possible inconsistencies. For the chemical module e.g. it is important to have a mass consistent advection and turbulence scheme which is not as important for a meteorological code. For the chemistry code, the wind fields have to be absolutely divergence free, but a wind field which is divergence free in one numerical formulation does not necessarily imply that it also keeps this quality in another numerical scheme. To prevent hazardous inconsistencies, meteorological and chemical codes therefore should be designed for each other. Recently, some fully coupled models have been presented (e.g. Liu et al.; 1997). The full coupling is probably the most flexible and accurate solution, because inconsistencies within the program code can automatically be excluded. Coupling between air pollution, meteorology and canopy exchange processes can be directly included in each direction. Metphomod (METeorology and PHOtochemistry MODel), the model presented here, is a comprehensive and fully coupled meteorological/chemical model for complex terrain. Its major design goals were: