Secondary inorganic aerosol simulations for Europe with special attention to nitrate

Abstract

Abstract.Nitrate is an important component of (secondary inorganic) fine aerosols in Europe. We present a model simulation for the year 1995 in which we account for the formation of secondary inorganic aerosols including ammonium sulphate and ammonium nitrate, a semi volatile component. For this purpose, the chemistry-transport model LOTOS was extended with a thermodynamic equilibrium module and additional relevant processes to account for secondary aerosol formation and deposition. During winter, fall and especially spring high nitrate levels are projected over north western, central and eastern Europe. During winter nitrate concentrations are highest in Italy, in accordance with observed data. In winter nitric acid, the precursor for aerosol nitrate is formed through heterogeneous reactions on the surface of aerosols. Modelled and observed sulphate concentrations show little seasonal variation. Compared to sulphate levels, appreciable ammonium nitrate concentrations in summer are limited to those areas with high ammonia emissions, e.g. the Netherlands, since high ammonia concentrations are necessary to stabilise this aerosol component at high temperatures. As a consequence of the strong seasonal variation in nitrate levels the AOD depth of nitrate over Europe is especially significant compared to that of sulphate in winter and spring when equal AOD values are calculated over large parts of Europe. Averaged over all stations the model reproduces the measured concentrations for NO 3 , SO 4 , NH 4 , TNO 3 (HNO 3 +NO 3 ), TNH 4 (NH 3 +NH 4 ) and SO 2 within 20%. The daily variation is captured well, albeit that the model does not always represent the amplitude of single events. The model underestimates wet deposition which was attributed to the crude representation of cloud processes.Comparison of retrieved and computed aerosol optical depth (AOD) showed that the model underestimates AOD signifiCorrespondence to: M. Schaap (m.schaap@mep.tno.nl) cantly, which was expected due to the lack of carbonaceous aerosols, sea salt and dust in the model. The treatment of ammonia was found to be a major source for uncertainties in the model representation of secondary aerosols. Also, inclusion of sea salt is necessary to properly assess the nitrate and nitric acid levels in marine areas.

PM measurement campaign HOVERT in the Greater Berlin area: model evaluation with chemically specified particulate matter observations for a one year period

“…Average differences for sulphate (over a set of over 40 EMEP sites) were in general within a range of ±20% in earlier evaluation studies for different European air quality models (Hass et al, 2003;Bessagnet et al, 2004;Schaap et al, 2004a), in accordance to our results for the HOVERT sites. However, average correlation is better in this work (0.72) than in previous studies (in the range 0.3-0.7).…”

“…Especially, filter techniques as used for this comparison are subject to evaporation of nitrate (and ammonium) above about 20 • C (e.g. Schaap et al, 2004a). Thus, part of the overestimation at rural sites during HOVERT (Paulinenaue +17%, Hasenholz +86%) and part of the overestimation encountered in previous studies (Hass et al, 2004) could be explained by measurement uncertainty.…”

“…In particular, evaporation of nitrate (and ammonium) above about 20 • C outside temperature (e.g. Schaap et al, 2004a) can lead to a substantial (but difficult to quantify) underestimation of these ions especially in summer. During an intercomparison campaign at a Dutch rural sites, total carbon measurements obtained with high volume samplers agreed within 10% (ten Brink et al, 2004).…”

“…This period is also related to enhanced observed and simulated photooxidant levels (O 3 +NO 2 ) pointing to an enhanced photochemical activity. Normalised root mean square errors (NRMSE) both reflect errors in the day to day variability and systematic biases (for a definition, see for example Schaap et al, 2004a;Hass et al, 2003). Typical values for HOVERT sites for a year period and nested simulations are about 50% for PM 10 , 60-70% for SO 2− 4 , 60-80% for NH + 4 , 80-110% for NO − 3 , 60-140% for EC and 50-70% for OC.…”

“…Air quality model evaluation studies using chemically specified PM data for secondary inorganic ions from the EMEP network are presented in a number of papers (Hass et al, 2003;Bessagnet et al, 2004;Schaap et al, 2004a;Tørseth et al, 2004). For most of the sites and most of the models, simulated sulphate, nitrate and ammonium summer and winter averages agree within a factor of two to observations.…”