Partner und Internationale Organisationen
(Englisch)
|
B, CH, CRO,CZ, D, DK, E, F, FIN, GR, H, I, N, S, SK, UK
|
Abstract
(Englisch)
|
The model Berphomod (Bernese photochemical model [1]) is an Eulerian model to simulate ozone and summer smog dynamics in time and space. Atmospheric flow physics, energy balance, turbulence, photochemistry, and exchange processes with terrestrial surfaces including dry deposition are integrated into one consistent execu-table program code. The model has been developed for simulations in topographically complex regions with a 50 - 200 km horizontal extension. Model runs usually consider periods of 2-7 days, the typical time scale for summer smog build-up. In the present study 2 areas were defined [2]. One region with a rectangular area of 120 km x 120 km was cen-tred over the city of Basel. The area was chosen to cover a large part of the area previously used for similar simulations done at IMK Karlsruhe with the model system KAMM/DRAIS for model comparisons. Pixel size is 2 km x 2 km, the grid is tilted by 1° counterclockwise. In the vertical dimension the model considers heights up to 3000 m, subdivided into 17 layers increasing in depth with increasing height. The other model region covers an area of 120 km x 140 km around Geneva. The area is rotated 38° counter-clock-wise from N-S-orientation to fit into the model grid used by the working group at EPFL Lausan-ne with the CIT model to allow easier generation of surface parameters and emission data. In the vertical dimension the model considers a volume up to 10 km height, divided into 20 layers increasing in depth with height. As a product of daytime photochemical processes, urban emissions lead to the formation of urban plumes where O3 concentration peaks are about 10-20 ppb higher compared to the surroundings. Simulated ozone concentrations generally show good agreement with measured values, particularly in the after-noon, when the surface layer is well mixed. At night, calculated values are often higher than the observed, mainly because measurements are done a few meters above ground, whereas modelled values represent volumes of typically 2 km x 2 km x 50 m. The simulated effects of emission reductions on surface ozone concentrations are consistent with results ob-tained in earlier studies. Combined reductions, concerning NOx and anthropogenic VOC emissions, appear to be most effective. In urban areas higher peak concentrations can occur, particularly if only NOx emissions are re-duced. If both precursor sources were reduced by 50 %, a decrease of around 15 ppb could be expected. Sensitivity analysis have revealed that the model domain should be chosen big enough in order to allow local winds to disperse the emissions realistically and to allow enough reaction time for photochemistry to take place. The accurate initialisation of trace gases has to be done very carefully. Alternatively a pre-run phase of one day must be allowed. [1] Perego S (1996) Ein numerisches Modell zur Simulation des Sommersmogs. PhD Thesis University of Berne. Geographica Bernensia G47, 202pp. [2] Künzle Th and Joss U (1997) Modelling Ozone Concentrations with Berphomod: Case Studies in Basel and Geneva. COST Action 615: Modelling of Urban Air Pollution. Edited by the Swiss Agency for the Environment, Forests and Landscape (SAEFL), Berne, 131 pp.
|
