Regional Air Quality (RAQ) is a measure as to how clean or polluted a local atmosphere is. The Acid Rain phenomenon of the late 1970’s motivated the application of abatement strategies targeting large-scale industrial processes and energy production. These are used to monitor the reduction of the emission of gaseous pollutants and their detrimental effects on buildings and eco-systems. Since the 1970’s the trend in Western Europe has generally been a decline in the size of manufacturing industry with a simultaneous rise in the use of motor vehicles. This has altered the chemical composition of the atmosphere by reducing sulphur containing compounds whilst increasing nitrogen containing compounds. A consequence of this is an increase in the in-situ formation of tropospheric ozone which is detrimental to human health and crop yields.

Medium-scale RAQ models are now commonly used for evaluating ‘extreme‘ episodes which occur. In recent years such models have become more reliable enabling them to be used in an operational system. The resulting simulations are then used to inform the public as to when high pollution levels are expected, such as those experienced during the European heatwave of 2003. As part of the EU GEMS project (http://gems.ecmwf.int/) a number of RAQ models are setup to deliver daily forecasts up to three days ahead on a routinely basis.

In order to validate this system a comparison of 7 of the RAQ models contributing to the project (namely BOLCHEM, CAC, CAMx, CHIMERE, EMEP, EURAD and MATCH) has been performed using the pre-operational forecasts for the second part of 2008 by Dr. V. Huijnen and Dr H. Eskes based at KNMI (KS-CK), in collaboration with a host of other research institutes. Three large-scale simulations using 3D Chemistry Transport Models (MOZART-IFS, TM5 and TM5-zoom) were also included and comparisons were made against NO2 columns retrieved from the Ozone Monitoring Instrument (OMI) for this period. The monthly mean retrieved for August 2008 from OMI is shown in Figure 1 below.

As the global models are designed to evaluate long-range transport events and large-scale features in the troposphere, the CTM’s did not have the ability to capture the daily cycle introduced by peaks in transport intensity (rush hours) as simulated in the RAQ models. On the other hand, the global models were generally well in line with the RAQ models, which validates the use of a combined system, where the global models deliver boundary conditions to the RAQ models.

Compared to OMI observations, the RAQ models captured the temporal differences relatively well although generally background levels were lower. This is most likely caused by a combination of model deficiencies and issues related to the retrieval algorithm.

The 3D CTM simulations showed that the resolution has an important effect on modeled [NO2], as the TM5-zoom model with a resolution of 1° x 1° improved its regional distribution. For the total tropospheric columns retrieved from OMI, there is a variation with season in line with the differences in the chemical lifetime of NO2 throughout the year. Comparing the RAQ simulations shows that there is a worse correlation for the summertime, especially for Eastern Europe.

These results have been presented at the Final GEMS assembly, which took place at Forschungzentrum Jülich between 31st March and the 3rd of April 2009, and will be presented at the forthcoming EGU General Assembly in Vienna between 19th-24th April, 2009

Poster presentation can be downloaded here (pdf 1.2 Mb).