Submicron ambient aerosol was characterized in summer 2005 at an urban background site in Zurich, Switzerland, during a three-week measurement campaign. Highly time-resolved samples of non-refractory aerosol components were analyzed with an Aerodyne aerosol mass spectrometer (AMS). Positive matrix factorization (PMF) was used for the first time for aerosol mass spectra to identify the main components of the total organic aerosol and their sources. The PMF retrieved factors were compared to measured reference mass spectra and were correlated with tracer species of the aerosol and gas phase measurements from collocated instruments. Six factors were found to explain virtually all variance in the data and could be assigned either to sources or to aerosol components such as oxygenated organic aerosol (OOA). Our analysis suggests that at the measurement site only a small (<10%) fraction of organic PM₁ originates from freshly emitted fossil fuel combustion. Other primary sources identified to be of similar or even higher importance are charbroiling (10–15%) and wood burning (~10%). The fraction of all identified primary sources is considered as primary organic aerosol (POA). This interpretation is supported by calculated ratios of the modelled POA and measured primary pollutants such as elemental carbon (EC), NO_x, and CO, which are in good agreement to literature values. A high fraction (60–69%) of the measured organic aerosol mass is OOA which is interpreted mostly as secondary organic aerosol (SOA). This oxygenated organic aerosol can be separated into a highly aged fraction, OOA I, (40–50%) with low volatility and a mass spectrum similar to fulvic acid, and a more volatile and probably less processed fraction, OOA II (on average 20%). This is the first publication of a multiple component analysis technique to AMS organic spectral data and also the first report of the OOA II component.

Meteorological normalisation is a technique which accounts for changes in meteorology over time in an air quality time series. Controlling for such changes helps support robust trend analysis because there is more certainty that the observed trends are due to changes in emissions or chemistry, not changes in meteorology. Predictive random forest models (RF; a decision tree machine learning technique) were grown for 31 air quality monitoring sites in Switzerland using surface meteorological, synoptic scale, boundary layer height, and time variables to explain daily PM₁₀ concentrations. The RF models were used to calculate meteorologically normalised trends which were formally tested and evaluated using the Theil–Sen estimator. Between 1997 and 2016, significantly decreasing normalised PM₁₀ trends ranged between −0.09 and −1.16 µg m⁻³ yr⁻¹ with urban traffic sites experiencing the greatest mean decrease in PM₁₀ concentrations at −0.77 µg m⁻³ yr⁻¹. Similar magnitudes have been reported for normalised PM₁₀ trends for earlier time periods in Switzerland which indicates PM₁₀ concentrations are continuing to decrease at similar rates as in the past. The ability for RF models to be interpreted was leveraged using partial dependence plots to explain the observed trends and relevant physical and chemical processes influencing PM₁₀ concentrations. Notably, two regimes were suggested by the models which cause elevated PM₁₀ concentrations in Switzerland: one related to poor dispersion conditions and a second resulting from high rates of secondary PM generation in deep, photochemically active boundary layers. The RF meteorological normalisation process was found to be robust, user friendly and simple to implement, and readily interpretable which suggests the technique could be useful in many air quality exploratory data analysis situations.

Equivalent black carbon (EBC) measured by a multi-wavelength Aethalometer can be apportioned to traffic and wood burning. The method is based on the differences in the dependence of aerosol absorption on the wavelength of light used to investigate the sample, parameterized by the source-specific absorption Ångström exponent (α). While the spectral dependence (defined as α values) of the traffic-related EBC light absorption is low, wood smoke particles feature enhanced light absorption in the blue and near ultraviolet. Source apportionment results using this methodology are hence strongly dependent on the α values assumed for both types of emissions: traffic α_TR, and wood burning α_WB. Most studies use a single α_TR and α_WB pair in the Aethalometer model, derived from previous work. However, an accurate determination of the source specific α values is currently lacking and in some recent publications the applicability of the Aethalometer model was questioned. Here we present an indirect methodology for the determination of α_WB and α_TR by comparing the source apportionment of EBC using the Aethalometer model with ¹⁴C measurements of the EC fraction on 16 to 40 h filter samples from several locations and campaigns across Switzerland during 2005–2012, mainly in winter. The data obtained at eight stations with different source characteristics also enabled the evaluation of the performance and the uncertainties of the Aethalometer model in different environments. The best combination of α_TR and α_WB (0.9 and 1.68, respectively) was obtained by fitting the Aethalometer model outputs (calculated with the absorption coefficients at 470 and 950 nm) against the fossil fraction of EC (EC_F ∕ EC) derived from ¹⁴C measurements. Aethalometer and ¹⁴C source apportionment results are well correlated (r = 0.81) and the fitting residuals exhibit only a minor positive bias of 1.6 % and an average precision of 9.3 %. This indicates that the Aethalometer model reproduces reasonably well the ¹⁴C results for all stations investigated in this study using our best estimate of a single α_WB and α_TR pair. Combining the EC, ¹⁴C, and Aethalometer measurements further allowed assessing the dependence of the mass absorption cross section (MAC) of EBC on its source. Results indicate no significant difference in MAC at 880 nm between EBC originating from traffic or wood-burning emissions. Using EC_F ∕ EC as reference and constant a priori selected α_TR values, α_WB was also calculated for each individual data point. No clear station-to-station or season-to-season differences in α_WB were observed, but α_TR and α_WB values are interdependent. For example, an increase in α_TR by 0.1 results in a decrease in α_WB by 0.1. The fitting residuals of different α_TR and α_WB combinations depend on EC_F ∕ EC such that a good agreement cannot be obtained over the entire EC_F ∕ EC range using other α pairs. Additional combinations of α_TR = 0.8, and 1.0 and α_WB = 1.8 and 1.6, respectively, are possible but only for EC_F ∕ EC between ∼ 40 and 85 %. Applying α values previously used in the literature such as α_WB of ∼ 2 or any α_WB in combination with α_TR = 1.1 to our data set results in large residuals. Therefore we recommend to use the best α combination as obtained here (α_TR = 0.9 and α_WB = 1.68) in future studies when no or only limited additional information like ¹⁴C measurements are available. However, these results were obtained for locations impacted by black carbon (BC) mainly from traffic consisting of a modern car fleet and residential wood combustion with well-constrained combustion efficiencies. For regions of the world with different combustion conditions, additional BC sources, or fuels used, further investigations are needed.

The identification of atmospheric trace species measurements that are representative of well-mixed background air masses is required for monitoring atmospheric composition change at background sites. We present a statistical method based on robust local regression that is well suited for the selection of background measurements and the estimation of associated baseline curves. The bootstrap technique is applied to calculate the uncertainty in the resulting baseline curve. The non-parametric nature of the proposed approach makes it a very flexible data filtering method. Application to carbon monoxide (CO) measured from 1996 to 2009 at the high-alpine site Jungfraujoch (Switzerland, 3580 m a.s.l.), and to measurements of 1,1-difluoroethane (HFC-152a) from Jungfraujoch (2000 to 2009) and Mace Head (Ireland, 1995 to 2009) demonstrates the feasibility and usefulness of the proposed approach. The determined average annual change of CO at Jungfraujoch for the 1996 to 2009 period as estimated from filtered annual mean CO concentrations is −2.2 ± 1.1 ppb yr⁻¹. For comparison, the linear trend of unfiltered CO measurements at Jungfraujoch for this time period is −2.9 ± 1.3 ppb yr⁻¹.

In March 2020, non-pharmaceutical intervention measures in the form of lockdowns were applied across Europe to urgently reduce the transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus which causes the COVID-19 disease. The aggressive curtailing of the European economy had widespread impacts on the atmospheric composition, particularly for nitrogen dioxide (NO₂) and ozone (O₃). To investigate these changes, we analyse data from 246 ambient air pollution monitoring sites in 102 urban areas and 34 countries in Europe between February and July 2020. Counterfactual, business-as-usual air quality time series are created using machine-learning models to account for natural weather variability. Across Europe, we estimate that NO₂ concentrations were 34 % and 32 % lower than expected for respective traffic and urban background locations, whereas O₃ was 30 % and 21 % higher (in the same respective environments) at the point of maximum restriction on mobility. To put the 2020 changes into context, average NO₂ trends since 2010 were calculated, and the changes experienced across European urban areas in 2020 was equivalent to 7.6 years of average NO₂ reduction (or concentrations which might be anticipated in 2028). Despite NO₂ concentrations decreasing by approximately a third, total oxidant (O_x) changed little, suggesting that the reductions in NO₂ were substituted by increases in O₃. The lockdown period demonstrated that the expected future reductions in NO₂ in European urban areas are likely to lead to widespread increases in urban O₃ pollution unless additional mitigation measures are introduced.

Low cost sensors for measuring atmospheric pollutants are experiencing an increase in popularity worldwide among practitioners, academia and environmental agencies, and a large amount of data by these devices are being delivered to the public. Notwithstanding their behaviour, performance and reliability are not yet fully investigated and understood. In the present study we investigate the medium term performance of a set of NO and NO₂ electrochemical sensors in Switzerland using three different regression algorithms within a field calibration approach. In order to mimic a realistic application of these devices, the sensors were initially co-located at a rural regulatory monitoring site for a 4-month calibration period, and subsequently deployed for 4 months at two distant regulatory urban sites in traffic and urban background conditions, where the performance of the calibration algorithms was explored. The applied algorithms were Multivariate Linear Regression, Support Vector Regression and Random Forest; these were tested, along with the sensors, in terms of generalisability, selectivity, drift, uncertainty, bias, noise and suitability for spatial mapping intra-urban pollution gradients with hourly resolution. Results from the deployment at the urban sites show a better performance of the non-linear algorithms (Support Vector Regression and Random Forest) achieving RMSE < 5 ppb, R² between 0.74 and 0.95 and MAE between 2 and 4 ppb. The combined use of both NO and NO₂ sensor output in the estimate of each pollutant showed some contribution by NO sensor to NO₂ estimate and vice-versa. All algorithms exhibited a drift ranging between 5 and 10 ppb for Random Forest and 15 ppb for Multivariate Linear Regression at the end of the deployment. The lowest concentration correctly estimated, with a 25 % relative expanded uncertainty, resulted in ca. 15–20 ppb and was provided by the non-linear algorithms. As an assessment for the suitability of the tested sensors for a targeted application, the probability of resolving hourly concentration difference in cities was investigated. It was found that NO concentration differences of 5–10 ppb (8–10 for NO₂) can reliably be detected (90 % confidence), depending on the air pollution level. The findings of this study, although derived from a specific sensor type and sensor model, are based on a flexible methodology and have extensive potential for exploring the performance of other low cost sensors, that are different in their target pollutant and sensing technology.