Chapter 4.12 Analyzing the response of a chemical transport model to emissions reductions utilizing various grid resolutions

Abstract

Within the framework of a project financed by the German EPA (Umweltbundesamt), photochemical and aerosol modelling simulations over Europe, Germany, the Federal State Brandenburg and the (agglomeration) metropolitan area of Berlin for the whole year 2002 were performed using the REM-CALGRID (RCG) model. The aim of the study was to analyze the relationships between regional and urban air quality levels and the modelled effectiveness of control measures in an urban area as a function of varying, horizontal grid resolution. The study should help to answer the question of how reliably regional scale model calculations can be used to compute air quality in urban areas. The RCG model was run at four different grid resolutions to assess the air quality in Berlin: 1. A European scale grid resolution of 0.25° Latitude by 0.5° Longitude (i.e., nominally about 30-km); 2. A national scale (Germany) grid having twice the resolution of grid #1 (i.e., approx. 15-km); 3. A Federal State (Brandenburg) grid having 8 times the resolution of grid #1 (i.e., approx. 4-km) and, 4. An urban (Berlin) grid having 32-times the resolution of grid #1 (i.e., approx. 1-km). High-resolution emissions data generated by the approach were available for Berlin and Brandenburg. These emissions were used for grids #3 and #4. The European and German scale calculations on grids #1 and #2, respectively, were carried out employing the top-down TNO emissions inventory for Europe for base-year 2000 at the 0.125° lat. by 0.25° longitude resolution of grid #2. Because the emissions estimates of the two inventories differ substantially for the greater Berlin area, the local bottom-up emissions were integrated into the large scale TNO data base, thus, allowing the usage of harmonized emissions at all scales. A base run for each scale was carried out and compared with PM10 and NO2 observations taken within the area of Berlin. It turned out that model performance was best for the run with the highest grid resolution. As might be expected, the urban increment was strongly underestimated by the regional (grid #1) model resolution. The results of two emissions scenarios, a 50% reduction of all emissions and the CLE-scenario (Implementation of current legislation until 2010), clearly showed that the effects of a measure will be underestimated in urban areas if they are based on a regional scale model application.