Spatio-temporal variability of levels and speciation of particulate matter across Spain in the CALIOPE modeling system

Abstract

The CALIOPE high-resolution air quality modeling system (4 km × 4 km, 1 h) estimates particulate matter from two aerosol models, CMAQv4.5 (AERO4) and BSC-DREAM8b. While CMAQv4.5 calculates biogenic, anthropogenic and sea-salt aerosols; BSC-DREAM8b provides hourly estimates of the natural mineral dust contribution from North Africa deserts. This paper presents an evaluation of the CALIOPE system to reproduce the spatial and temporal variability levels of PM2.5, PM10 and chemical composition (nitrate, non-marine sulfate, ammonium, organic and elemental carbon, sea-salt, and desert dust) across Spain. The evaluation is performed against ground-based observations for the year 2004, when a number of time series of chemically speciated compounds were available. A new data set of Saharan dust PM10 concentration is used to evaluate the PM10 contribution modeled by BSC-DREAM8b. The results indicate that both natural aerosol sea-salt and desert dust accomplish the model performance criteria (MFE ≤ 75% and MFB ± 60%). Modeled PM10 sea-salt is highly dependent on wind speed and presents high correlation with experimental data in coastal areas (r = 0.67). The BSC-DREAM8b is able to reproduce the daily variability of the observed levels of desert dust and most of the outbreaks affecting southern Spain. Species in the equilibrium (e.g. sulfate/nitrate/ammonium) are highly correlated each other and show high dependency on ammonia emissions. Non-marine sulfate and ammonium are underestimated by a factor of 3. An underestimation of nitrate was also seen (factor of 2). Fine carbonaceous aerosols present the highest underestimations (factor of 4) in part related to the state-of-the-science concerning secondary organic aerosol formation pathways. Spatial and seasonal variability of PM2.5, PM10 and their chemical compounds increase the correlation with observations when multiplicative bias-correction factors for the aforementioned underestimated species are taking into account. Furthermore, simulated spatial and seasonal patterns of aerosol agree with those described in related studies based on experimental values.