J. Pauraitė a, I. Garbarienė a, A. Minderytė a, V. Dudoitis a, G. Mainelis b, L. Davulienė a, I. Uogintė a, K. Plauškaitė a, and S. Byčenkienė a

a Department of Environmental Research, SRI Center for Physical Sciences and Technology, Savanorių 231, 02300 Vilnius, Lithuania

b Rutgers, The State University of New Jersey, 14 College Farm Road, New Brunswick, NJ 08901, United States
Email: julija.pauraite@ftmc.lt

Received 20 January 2021; revised 15 April 2021, accepted 16 April 2021

Open biomass burning (OBB) is a significant air pollution source, but it is still not clear to what extent OBB events affect indoor air quality [1]. Outdoor and indoor measurements of submicron particulate matter (PM1) were conducted on 25–29 April (2019) in the capital city Vilnius (Lithuania). Fires from neighbouring countries (Belarus, Ukraine and Russia) and in the vicinity of Vilnius broke out during the measurement campaign. The temporal evolution and transport of OBB plume were investigated by combining the air mass backward trajectory analysis and fire satellite observation (MODIS) database. Measurements of the PM1 chemical composition in real-time were performed using an aerosol chemical speciation monitor (ACSM) and an aethalometer. Organic matter was the clearly dominant component, accounting for >70%, in both indoor and outdoor PM1. The air filtering system of the office building removed approximately up to 55% of PM1. Despite a significantly lower PM1 pollution level in the office, highly acidic indoor PM1 could have harmful effects on the human health. Source apportionment of particulate carbonaceous matter revealed a significant importance of OBB-related particles (average 56%) to indoor air.

Keywords: organic aerosol, black carbon, indoor air quality, biomass burning

PACS: 92.60.Sz, 92.60.Mt, 92.60.hf

1. Introduction

Spring grass fires (or open biomass burning – OBB) is a  significant air pollution source, with adverse impacts on global, regional and local air quality, public health and climate [25]. OBB occurs globally with highly variable biomass types (wood, shrub, grass, vegetation and crop residues) and burning conditions [6]. OBB generally emits a high amount of gaseous compounds (CO2, CO, NOx, volatile organic compounds – VOCs) and carbonaceous aerosols (e.g. equivalent black carbon (eBC) and organic carbon (OC)) to the  atmosphere in a short period of time [7, 8]. However, it was estimated that OBB globally contributed to ~40% of the  annual average submicron eBC emission and ~65% of primary OC emission [3]. Quantitative assessment of the contribution of biomass burning to carbonaceous aerosol has focused on the molecular markers, such as levoglucosan, approach [9, 10]. Recently, the characterization of chemical composition and potential sources of submicron aerosol particles (PM1) has improved with the  advent of aerosol mass spectrometric technique (AMS, aerosol chemical speciation monitor – ACSM). Previous studies demonstrated that the application of factor analysis approach on ACSM derived mass spectra of organic species permits the identification of various organic aerosol (OA) sources and photochemical ageing processes [11, 12]. The source apportionment model based on light absorption properties of carbonaceous materials (aethalometer model) has been widely used to separate combustion (fossil fuel and biomass burning) and non-combustion sources (secondary organic aerosol) of carbonaceous aerosol [13, 14].

Lithuania and surrounding countries are frequently affected by high-pollution events in early spring due to illegal grass burning for land clearing [15, 16]. Long-term measurements of carbonaceous aerosols revealed the yearly occurrence of intense biomass/grass burning events during March–April related to the  regional aerosol transport from the  Kaliningrad region, Ukraine and the  southwestern part of Russia surrounding the Black Sea [1518]. Moreover, forest and grassland burning phenomena could significantly influence indoor pollution levels [19].

The quality of indoor air has become a  global concern during last decades as people spend most part of the day inside of buildings [20, 21]. Studies indicate that the indoor air quality is influenced by the outdoor air quality mainly due to the building ventilation factor [21, 23]. Modern energy-conserving buildings have improved energy conservation through reducing exchanges between outdoor and indoor air, and synthetic materials and chemical products have been extensively used in these airtight buildings. Therefore, the selection of ventilation regimes and monitoring of indoor sources become important [24, 25]. Numerous studies focused on indoor air quality within residential buildings [23, 2628]. However, the  impact of short-term air pollution events caused by open biomass/grass burning on the  office building microenvironment has not been extensively investigated.

The aim of this study was to evaluate the influence of local/regional open biomass burning events on indoor air quality and sources distribution in a mechanically ventilated office building.

2. Materials and methods

2.1. Measurement site

Complex measurements were performed on 25–26 April in indoor air and on 26–29 April 2019 in outdoor air. The  sampling site was ~370  km northwest from active fire regions. The investigation was carried out at a newly built (2015) lowenergy (B class) building of the Center for Physical Sciences and Technology in Vilnius, Lithuania (54º43′24′′N, 25º19′36′′E). The  instruments were placed in the  vacant meeting room on the  2nd floor. The  door of the  meeting room was kept closed to avoid air exchange with other rooms and minimize the impact of indoor PM1 sources. During the  outdoor measurement period inlets were positioned outside. The  mechanical/forced air ventilation system in this room is equipped with G4/F7 filters. The pre-filter G4 removes particles in the  >10  µm size range. The  F7 filter removes particles within 0.3–10 µm. The rate of air flow into and out of the  meeting room was 320 and 120–320 mh–1 (depending on CO2 amount), respectively. The sampling site is located 6–7 km northeast from the  centre of Vilnius. The  local traffic is usually an important source of PM1 in Vilnius (25%) [29]. Because of the heating season termination 10 days earlier, during this period the influence of domestic heating was negligible. The biomass burning (BB) source could be related to OBB (e.g. grassland burning, forest and garden waste burning) that routinely shows a significant contribution during spring in Lithuania.

2.2. Instrumentation

Equivalent black carbon (eBC) measurements were performed using an aethalometer (Model AE31 Spectrum, manufactured by Optotek, Slovenia). The  optical transmission of carbonaceous aerosol particles was measured sequentially at seven wavelengths λ (0.37, 0.45, 0.52, 0.59, 0.66, 0.88 and 0.95 µm) with 2 min intervals and 3.9 l/min flow rate. The aethalometer model [13] was applied in order to attribute eBC mass concentration to biomass burning and fossil fuel combustion (eBCBB and eBCFF, respectively). Filter-based optical measurements contain systematic errors due to filter loading, ‘shadowing’, multiple scattering and other effects [30]. Thus, the  correction suggested by Weingartner was applied in this study; this method is described in more detail in the original article by the author [31].

During this study, the concentration of organic aerosol (OA) together with secondary inorganic aerosol (SIA) was measured by an aerosol chemical speciation monitor ACSM (Aerodyne Research, Inc., Billerica, MA, USA) which operated with a time resolution of ~30 min. A detailed description of the ACSM operation and data analysis are given in Ref. [17], Ulevicius et al. (2016).

2.3. Models and additional tools

Due to short both indoor and outdoor measurement campaigns, positive matrix factorization analysis could not be statistically robust. Therefore, an alternative source apportionment was implemented. The analysis consisted of the assessment of different carbonaceous species (CM) and OA mass spectra attribution for each of them.

The aethalometer model was originally developed in order to quantify eBC sources (fossil fuel and biomass burning). However, a  few studies adapted an aethalometer model to quantitatively estimate the  contribution of particulate carbonaceous matter (eBC and organic mass (OM)) from fossil fuel and biomass burning emissions, using the  different light absorption parameters of these sources [3235]. We applied this approach to apportion the total carbonaceous material (CM) mass concentration to biomass burning (CMBB), fossil fuel (CMFF) and non-absorbing material (CMNA) according to the following equations:

CM= CM FF + CM BB + CM NA ,(2)
CM= c 1 b absFF (950nm)+ c 2 b absBB (470nm)+ c 3 .(3)

In our study, the  total CM is the  sum of eBC, measured by the aethalometer, and OM measured by ACSM. The  CM mass concentration was regressed against babsFF (950 nm) and babsBB (470 nm) for the  estimation of contribution of transport emissions (CMFF) and biomass burning (CMBB) sources. The  coefficients c1 and c2, calculated by Eq. 3, are related to the light absorbing CM mass of both sources. The intercept c3 reflects the contribution of non-absorbing carbonaceous materials (CMNA).

In our study, the mass spectra of different CM species were estimated by attributing the time series of OA to the related m/z signals profiles. Based on bilinear modelling [11], OA can be presented as the product of two matrices. The first one comprises the  time series of mass concentration (TS) while the second one is the mass spectrum (MS) or source profiles of individual OA factors


where (E) is the matrix of residuals. In our study, the time series of CMBB, CMFF and CMNA were used as individual TS [11]:

OA ij = p=1 p TS ip MS pj + E ij ,(5)

Here j refers to an ion fragment at time step i for a given factor p. As a result, individual MS of OA for each CM species were obtained. The ME-2 engine tool [36] was used for the calculation.

The mass spectra of outdoor and indoor identified factors have been compared using the θ angle analysis introduced by Kostenidou et al. (2009) [37]. The θ angle is a measure of similarity between two mass spectra (θ < 15° high similarity; 15° < θ < 30° partial similarity; θ > 30° different spectra).

Wildfire events were explored using the fire information for the  resource management system (FIRMS) which distributes satellite observation from the  NASA’s moderate resolution imageing spectroradiometer (MODIS) and the visible infrared imaging radiometer suite (VIIRS).

Air mass backward trajectories were calculated using the  hybrid single-particle Lagrangian integrated trajectory (HYSPLIT4) [38] model with the global data assimilation system (GDAS) meteorological databases at the NOAA Air Resources Laboratory’s web server (NOAA HYSPLIT Trajectory Model. Ready.Noaa.Gov, 2020, https://www.ready.noaa.gov/hypub-bin/trajtype.pl?runtype=archive). To provide a better view of which air masses had influenced the rise of aerosol concentrations, trajectory frequencies were calculated using the NOAA HYSPLIT model, where the 72 h backward trajectory frequency was calculated from a single location (sampling site) at 500 m height every 6 h. The sum of frequency of trajectories passed over a grid cell (1.0° × 1.0°) was normalized by the total number of trajectories [39].

In our study, aerosol pH was evaluated by a  thermodynamic equilibrium model ISORROPIA-II [40]. This method calculates the  equilibrium partitioning the  given total concentration of different species. ISORROPIA-II determines the  system of equilibrium equations and solves them for the  equilibrium state using the  chemical potential method [40]. pH was evaluated using aerosol secondary inorganic chemical components (NH4+, NO3, SO42–, Cl) measured by ACSM together with temperature (T) and relative humidity (RH).

RH, T and time series of NO2 were measured at the nearby (2 km away) located monitoring station.

2.4. Meteorological condition during the study period

The average daytime temperature on 25–29 April was about +22°C (positive 1.2–2.9° anomaly). The  highest daytime temperature was fixed at +29°C. The  average night-time temperature was +10–12°C. The  direction of the  wind was mainly from east during the whole study period. The wind speed was relatively low and varied between 1 and 7 m s–1. The air pressure ranged from 1040 to 1200 hPa. 2019 April was one of the driest months in the last decade. During the day, most of the time slightly or moderately unstable conditions with a boundary layer depth up to 1.5–2.0 km were dominant. On 25–29 April, in most of Lithuania there was no precipitation at all or the amount of precipitation did not reach 5.0 mm (0.0–0.1 standard precipitation rate – SPR). Due to the abovementioned meteorological conditions, the high fire risk (Level 4) was registered in all Lithuanian municipalities. On 25–29 April 2019, there were 266 (about 125 ha burnt area) fires in the open territory (forest, grass, etc.) in Lithuania. 100 of them (about 52.5 ha) were in the  Vilnius District (http://pagd.lrv.lt/lt/paros-ivykiai/2019-m-1/2019-m-balandis, last visited on 13 April 2021). The following indoor meteorological conditions were fixed: RH = 40% and T = 21°C.

2.5. Fire event

The active fire map created using the  FIRMS MODIS database confirms the great abundance of active fire events in the regional area eastwards from Vilnius for the  period of 25–29 April (Fig.  1(a)). The highest density of the active fire locations was present in the Kaliningrad Region of Russia that is 300–400 km to east from the measurement station. Meanwhile, another OBB located near the border of Belarus and Ukraine was ~370  km southeast from Vilnius.

Figure  1(b) shows the  smoke surface concentration measured on 25  April. It is worth noting that the smoke concentration was enhanced due to large quantities of active fires as the highest concentrations are observed over the regions with denser locations of active fires. Furthermore, Fig. 1(c) presents a  backward 72  h trajectory frequency plot where the  arrival of air parcels originating aloft over wildfire regions and other regions with a high number of active fire locations is apparent. Thus, the influence of smoke transport for surface-level air quality in Vilnius is evident. On 25  April, air parcels crossed the  central part of Ukraine and then southwest of Belarus which demonstrates that the air parcels passed over the OBB region and other active fire locations and later descended to the surface level in the  area of Vilnius where enhanced PM levels were recorded. The  backward trajectories of air parcels arriving to Vilnius show a very similar path on 25–27 April where the air masses originate aloft presence of wildfires. The mass concentration of particulate matter with a diameter up to 10 µm (PM10) in the capital city Vilnius (Lithuania) rose up to 90 µg m−3 on 25–27 April. The daily PM10 concentration exceeded the European Union’s established limit value (50 µg m−3) up to 1.6 times (http://oras.gamta.lt/files/Oras_190426v.pdf, last visited on 13  April 2021). Later on 28–29 April, the  air mass transport from northeast prevailed. During the following period, the density of active fire was significantly lower, resulting in the  decrease of PM10 concentration (average 30 µg m−3). However, this PM10 level exceeded the typical nonepisodic condition in Vilnius (16.8±8.4  µg  m−3) (average  ±  standard deviation). Consequently, it can be concluded that air parcels from the southeast region had a significant influence on enhanced pollutant concentrations in Vilnius on 25–29 April.

Fig. 1. Locations of active fire detections from the FIRMS MODIS database for the time period from 25 to 29 April (a), NRL model results showing smoke surface concentrations for 25 April (b), NOAA HYSPLIT model backward 72 h trajectory frequency plot from a starting location at the measurement station in Vilnius at 500 m AGL (c).

We used the records of cloud-aerosol lidar and infrared pathfinder satellite observations (CALIPSO) to investigate the sources of aerosol particles and the  vertical distribution of the  25  April 2019 pollution episode in Vilnius. The satellite observation from CALIPSO showed (Fig.  2) that aerosol load below 3 km was classified as polluted dust, smoke and polluted continental. These findings indicated that the  regional/local biomass burning events were most likely the cause of enhanced pollution levels in Vilnius, Lithuania.

3. Results

3.1. PM1 variation, chemical composition and acidity

The time series of PM10 and PM1 chemical composition (organic compounds, sulphate, nitrate, ammonium and black carbon) and their relative contributions are presented in Fig. 3. The average mass concentration of PM1 of outdoor measurements was 43.2±22.3 µg m–3 while the average PM1 concentration of indoor air reached 8.8±2.7  µg  m–3. The concentration of PM1 (in both outdoor and indoor measurements) was strongly correlating with PM10 (measured only outdoors), r = 0.7, indicating a  substantial contribution of outdoor sources to PM1 levels in office buildings. In order to evaluate the  indoor air filtering, the  PM1 dataset was compared to the outdoor PM10 measurements. For the outdoor dataset, PM1 showed a significant contribution to PM10 (58±23%) while the indoor PM1 was 26±6% of the outdoor PM10. Thus, the air filtering system of the building removed approximately up to 55% of PM1.

The outdoor PM1 mass concentration was dominated by organic compounds (37.9±19.2 µg m–3), followed by SO42– (2.2±0.5  µgm–3), NO3 (2.1±1.3 µg m–3), eBC (0.9±0.6 µg m–3) and NH4+ (0.8±0.3  µg  m–3) (Fig.  3(a)). During the  intense OBB event (25–26 April), a  significant enhancement of organic compounds (up to 67.7 µg  m–3), NO3 (4.1 µg m–3) and eBC (3.1 µg m–3) mass concentration was registered. The  average mass contribution of PM1 species during the  OBB event was organic aerosol (87%), NO3 (5%), SO42– (5%), NH4+ (2%) and eBC (2%) (Fig. 3(b)).

The average mass concentrations of PM1 chemical components in office air were significantly lower (7.0±2.2 µg m–3 for organics, 1.1±0.3 µg m–3 for sulphate, 0.3±0.2 µg m–3 nitrate, 0.3±0.1 µg m–3 ammonium and 0.2±0.1 µg m–3 eBC). However, the chemical composition of PM1 was similar in both outdoor and indoor environments (Fig. 3). The highest difference in the  PM1 composition between outdoor and indoor measurements was observed for SO42– contribution (5 and 13%, respectively).

Fig. 2. CALIPSO-derived vertical profile of aerosol subtypes over Vilnius on 25 April 2019.
Fig. 3. Time series of PM1 species (a) and their contributions to total PM1 (b) for outdoor and indoor measurements. Time series of PM10 (a) was measured only outdoor.

Such enhancement in the  contribution of sulphate in the office building resulted in changes of PM1 acidity. As seen in Fig. 4, the average aerosol pH of indoor aerosol was significantly lower than the outdoor one. In our study, in both outdoor and indoor environments, aerosol pH negatively correlated with SO42– (r = 0.63 and r = 0.62, respectively). This is in agreement with previous studies [41, 42], which have shown that under low aerosol liquid water content conditions the  increase in SO42– levels was followed by decreased aerosol pH. Furthermore, the  same studies showed that SO42– had a greater effect on aerosol pH than NO3. Meantime in our study, only in the outdoor measurements a possible link between aerosol pH and NO3 was observed (r = 0.82) while no such correlation was observed in the indoor measurements. A similar pattern was observed between the aerosol pH and NH4+ concentration. Chen et al. (2019) [41] concluded that increased NH4+ resulted in a slight increase of aerosol pH. In agreement with the latter, in our study of outdoor measurements, the closest to neutral aerosol pH (pH 4.0–4.8) was observed with the  highest NH4+ concentrations. The  same link was not observed for the  indoor dataset (Fig. 4(b)).

Fig. 4. Aerosol pH versus SO42– mass concentration for outdoor (a) and indoor (b) measurements. Colour (online) plotting by NH4+ mass concentration.

Differences between indoor and outdoor aerosol pH could be caused by several factors. First of all, while outside the building the T and RH were on the average 16.8±4.2°C and 44±9%, respectively, the  indoor air conditions were set to be constant with higher T (21ºC) and significantly lower RH (25%). Ding et al. (2019) [42] showed that there was a significant decrease in pH with increased T, while the effect due to RH varied over different seasons. Meantime, the modelling results [41] revealed that warmer and dryer atmosphere led to lower aerosol pH because partitioning of ammonia to the  condensed phase is less favoured. In addition, other parameters such as size distribution and aerosol liquid water content could play a significant role in aerosol pH alterations [42]. Guo et al. (2015) [43] underlined that low pH could affect the solubility of metals such as Fe and Cu which might increase the toxicity of redox metals. Furthermore, a recent study [41] underlined that pH ranging from 0 to 3 could be called ‘sensitive window’ where nitrate related gas-to-particle partitioning is sensitive to pH alterations. Thus, low levels of indoor aerosol pH could cause differences in particle formation processes and cause alterations in other parameters, such as toxicity. Despite the significantly lower PM1 and concentrations of its chemical components in the office, highly acidic indoor PM1 could have harmful effects on the  human health. Therefore, additional study of various parameters of indoor air is needed.

3.2. Ageing properties of organic aerosol

Organic compounds were the most important constituents of PM1, with a  fraction of ~80% during the investigated period. The triangle plot [12] was used as a tool to explore possible sources and to assess OA atmospheric ageing properties. The triangle plot consists of f43 (intensity of m/z 43 normalized to OA) and f44 (intensity of m/z 44 normalized to OA) space with additional lines for the visualization (Fig.  5). Since photochemical ageing leads to the increase in m/z 44 fraction, the f44 axis can be considered as a  proxy of ageing parameterization. Several studies showed that different positions in the  triangle can be connected to specific OA sources [12, 44, 45]. The triangle space with f43 vs f44 range 0.03–0.07 and 0.10–0.14, respectively, could be related to long-range transport (LRT) biomass burning OA [45]. The area with the same f43 range but lower f44 values was assigned to local and fresh biomass burning OA while the  area with higher f44 was linked to aged low-volatile OA [45]. In our study, most of outdoor data points were located at the  area assigned to LRT biomass burning OA proving the previous hypothesis of a strong influence of OBB events. Several points were lower indicating the fresh biomass burning related emissions. 92% of the outdoor data points had f60 (intensity of m/z 60 normalized to OA) above the background level (no significant presence of the main signal of biomass burning tracer levoglucosan) of 0.003 [46], indicating that a great part of OA in Vilnius can be related to OBB emissions (OA-OBB). The average outdoor f60 values of OA-OBB were 0.005±0.001, additionally supporting that OA related to biomass burning from distance sources (LRT OA-OBB) exceeded those of local (fresh) biomass burning emissions.

Particles with even more advanced ageing processes were observed in the indoor measurements (f44 up to 0.25) (Fig. 5). Only 26% of indoor OA had f60 above 0.003 and could be attributed to biomass burning related OA (OA-BB). The  mean indoor f60 values (0.002±0.001) of OA-BB were close to the background level. These could be explained by the dilution effect when the OBB plume is mixed with air masses containing organic aerosol particles of other origin. On the other hand, the gradual decay of f60 could return f60 to background values [47]. Recent studies highlighted that aged OA-BB had a weak signal of m/z 60, suggesting that f60 is not an effective tracer of aged OA-BB [44, 48].

Fig. 5. Triangle plot of f43 versus f44 coloured (online) by f60 for outdoor (a) and indoor (b) measurements.

3.3. Apportionment of carbonaceous aerosol sources

The apportionment of eBC sources based on the aethalometer model showed that the eBCBB contribution to the eBC mass concentration was dominant for both outdoor and indoor measurements (68 and 67%, respectively) (Fig. 6). eBCFF made up only 32–33% of eBC mass during the  investigated period. The time series of eBCBB correlated best with f60 (r = 0.66). Furthermore, the intensity of signal f60 was higher than 0.003. Therefore, increased outdoor levels of eBCBB were associated with OBB events. Meanwhile, eBCFF showed a moderate correlation with NO2 in the gas phase (r  =  0.54) strengthening the hypothesis of possible origin from fossil fuel combustion. In addition, the diurnal plot showed that the highest values of eBCFF were reached during morning rush hours (from 7 to 10 am). Thus, these observations allowed to link eBCFF to the traffic related emissions (Fig.  6). The  previous study of eBC in Vilnius during the warm season under normal conditions showed a  dominant influence of eBCFF (92%) [49]. That finding suggested that the PM1 chemical composition on 25–29 May 2019 was clearly unusual for Vilnius and was related with the OBB event. In our study, the contribution of eBCFF and eBCBB to the total eBC mass concentration remained unchanged in outdoor and indoor environments. Thus, both eBCFF and eBCBB were filtered in the indoor air with the same efficiency and therefore a high influence of OBB related eBCBB remained significant.

In the  case of CM, biomass burning (CMBB) and non-absorbing carbonaceous materials (CMNA) were found to account, respectively, for 87 and 13% of the total CM during the outdoor measurements. The  average relative contribution of indoor CMBB, CMFF and CMNA was 56, 6 and 38%, respectively. The  results of the  model revealed a significant contribution of absorbing organic matter or brown carbon (BrC) to the  CM mass concentration.

Fig. 6. Time series of the eBCBB and eBCFF mass concentrations (a) and the diurnal cycles (b) of eBCBB, eBCFF, f60 and NO2 during the OBB event.

During the measurement campaign CMFF contribution to the  total CM was the  lowest (0% in outdoor and 6% (0.47±0.29 µg m–3) in indoor air). Due to low concentrations of CMFF and no meaningful or robust aerosol mass spectra were attributed to this aerosol. The time series of CMBB correlated well with eBCBB (= 0.85) and f60 (= 0.71), suggesting that the  CM source apportionment method is valid for our study and is an appropriate method for characterizing BB emissions. Furthermore, the  CMBB mass spectrum showed characteristics of both OA-BB (m/z 60) and oxygenated OA (m/z 18 and 44) (Fig. 7). This factor represents aged OA-BB or a  mixture of primary OA-BB and secondary OA from other sources. It is possible that the traffic derived secondary OA can partially contribute to CMBB.

Fig. 7. Time series of each CM component obtained using an advanced aethalometer model (a) and its mass spectra for outdoor and indoor measurements (b).

However, the  indoor data showed a  negligible signal of f60 in the CMBB profile (0.002±0.001) (Fig. 7). Moreover, a poor correlation between f60 and CMBB was observed in the  indoor samples, likely due to the  photochemical degradation of levoglucosan [50]. Recent studies [44, 49] demonstrate that the  mass spectral fingerprint of levoglucosan (f60) cannot be used as BB tracer in aged BB emissions (>1 day). These findings suggest that the  method of CM source apportionment based on optical properties is indeed a good BB indicator. Hence, the advanced aethalometer model was useful for the CM source apportionment of aged BB aerosol.

The CMNA component, represented by the highest contribution of f44 (18 and 25% in outdoor and indoor air, respectively), was attributed to the processed OA. This is associated with the  chemical ageing of OA during long-range transported OA.

The theta angle analysis showed that outdoor and indoor CMBB mass spectra were slightly different (θ = 16°). This is consistent with the chemical ageing of OA-OBB indoors. The  CMNA mass spectra in outdoor and indoor air were quite similar (θ = 11°). Many similarities of the outdoor and indoor spectra of both CMBB and CMNA profiles indicated that carbonaceous particles originating outdoors were an important contributor to the indoor CM mass concentration.

These results demonstrate that even if the indoor air had lower CM levels, the  influence of outdoor air was evident and therefore a significant importance of OBB related particles to indoor air was observed.

4. Conclusions

During the measurement campaign (25–29 April 2019), open biomass burning OBB occurred in the vicinity of Vilnius (Lithuania) and surrounding areas. The analysis of air mass backward trajectories and the results of satellite fire observation showed that the elevated concentration of PM1 in Vilnius was related to large quantities of active fires in Belarus and Ukraine. During the  whole measurement campaign, PM1 mass concentration was dominated by organic aerosol OA in both outdoor and indoor environments the mass concentration of which on the  average decreased from 37.9 to 7.0 µg m–3, respectively.

The pH of PM1 ranged between 2–4 and 1.7–2.5 outdoor and indoor, respectively. Lower pH was detected in PM1 with a higher load of SO42–. The aerosol pH analysis showed that due to specific meteorological conditions (low RH, 25%, and high T, 21°C) of the office air, pH levels were significantly lower and therefore could have led to changes in several PM processes and properties, including particle composition, reactivity, gas-particle partitioning and toxicity. Our observations revealed that the  air filtering system of the  building removed approximately up to 55% of PM1.

The variation of f44, f43 and f60 parameters allows us to discriminate BB from other sources and examine the  evolution of biomass burning organic aerosol OA-BB. At the beginning of the OBB episode, long-range transported BB emissions contributed significantly to OA. On 27–29 April, organic aerosol was characterized by high f44 and low f60 values, due to the oxidation processes and ageing of BB plume. Due to the OBB event, biomass burning derived carbonaceous aerosol was dominant for both outdoor (eBCBB = 68%, CMBB = 87%) and indoor (eBCBB = 67%, CMBB = 56%) measurements. It can be concluded that the  OBB event had a strong influence on the indoor air quality. Yet, additional studies of simultaneous indoor and outdoor air quality over OBB events are required before firm conclusions could be drawn.


This research was funded by Grant No.  S-MIP-20-28 ‘Contribution of Outdoor Air Pollution to Indoor Air Quality and Inhaled Deposited Dose’ from the  Research Council of Lithuania. The  authors also want to express their appreciation to the Environmental Protection Agency of Lithuania.


[1] P.  Rajagopalan and N.  Goodman, Improving the indoor air quality of residential buildings during bushfire smoke events, Climate 9, 32 (2021), https://doi.org/10.3390/cli9020032

[2] A. Kocbach Bølling, J. Pagels, K.E. Yttri, L. Barregard, G. Sallsten, P.E. Schwarze, and C. Boman, Health effects of residential wood smoke particles: the  importance of combustion conditions and physicochemical particle properties, Part. Fibre Toxicol. 6(1), 29 (2009), https://doi.org/10.1186/1743-8977-6-29

[3] T.C. Bond, S.J. Doherty, D.W. Fahey, P.M. Forster, T. Berntsen, B.J. DeAngelo, M.G. Flanner, S. Ghan, B. Kärcher, D. Koch, et al., Bounding the role of black carbon in the climate system: A scientific assessment, J. Geophys. Res. Atmos. 118(11), 5380–5552 (2013), https://doi.org/10.1002/jgrd.50171

[4] Ö. Gustafsson, M. Kruså, Z. Zencak, R.J. Sheesley, L. Granat, E. Engström, P.S. Praveen, P.S.P. Rao, C. Leck, and H. Rodhe, Brown clouds over South Asia: biomass or fossil fuel combustion? Science 323(5913), 495 (2009), https://doi.org/10.1126/science.1164857

[5] J.K. Zhang, M.T. Cheng, D.S. Ji, Z.R. Liu, B. Hu, Y. Sun, and Y.S. Wang, Characterization of submicron particles during biomass burning and coal combustion periods in Beijing, China, Sci. Total Environ. 562, 812–821 (2016), https://doi.org/10.1016/j.scitotenv.2016.04.015

[6] W.  Knorr, F.  Dentener, J.-F.  Lamarque, L.  Jiang, and A.  Arneth, Wildfire air pollution hazard during the 21st century, Atmos. Chem. Phys. 17, 9223–9236 (2017), https://doi.org/10.5194/acp-17-9223-2017

[7] Y. Zhang, D. Obrist, B. Zielinska, and A. Gertler, Particulate emissions from different types of biomass burning, Atmos. Environ. 72, 27–35 (2013), https://doi.org/https://doi.org/10.1016/j.atmosenv.2013.02.026

[8] R.J. Yokelson, J.D. Crounse, P.F. DeCarlo, T. Karl, S. Urbanski, E. Atlas, T. Campos, Y. Shinozuka, V. Kapustin, A.D. Clarke, et al., Emissions from biomass burning in the Yucatan, Atmos. Chem. Phys. 9(15), 5785–5812 (2009), https://doi.org/10.5194/acp-9-5785-2009

[9] A. Piazzalunga, C. Belis, V. Bernardoni, O. Cazzuli, P. Fermo, G. Valli, and R. Vecchi, Estimates of wood burning contribution to PM by the macro-tracer method using tailored emission factors, Atmos. Environ. 45(37), 6642–6649 (2011), https://doi.org/https://doi.org/10.1016/j.atmosenv.2011.09.008

[10] C. Perrino, L. Tofful, S.D. Torre, T. Sargolini, and S.  Canepari, Biomass burning contribution to PM10 concentration in Rome (Italy): Seasonal, daily and two-hourly variations, Chemosphere 222, 839–848 (2019), https://doi.org/10.1016/j.chemosphere.2019.02.019

[11] Q.  Zhang, J.L.  Jimenez, M.R.  Canagaratna, I.M. Ulbrich, N.L. Ng, D.R. Worsnop, and Y. Sun, Understanding atmospheric organic aerosols via factor analysis of aerosol mass spectrometry: a  review, Anal. Bioanal. Chem. 401(10), 3045–3067 (2011), https://doi.org/10.1007/s00216-011-5355-y

[12] N.L. Ng, M.R. Canagaratna, Q. Zhang, J.L. Jimenez, J. Tian, I.M. Ulbrich, J.H. Kroll, K.S. Docherty, P.S. Chhabra, R. Bahreini, et al., Organic aerosol components observed in Northern Hemispheric datasets from Aerosol Mass Spectrometry, Atmos. Chem. Phys. 10(10), 4625–4641 (2010), https://doi.org/10.5194/acp-10-4625-2010

[13] J.  Sandradewi, A.S.H.  Prévôt, E.  Weingartner, R. Schmidhauser, M. Gysel, and U. Baltensperger, A  study of wood burning and traffic aerosols in an Alpine valley using a  multi-wave-length Aethalometer, Atmos. Environ. 42(1), 101–112 (2008), https://doi.org/10.1016/j.atmosenv.2007.09.034

[14] E.  Liakakou, I.  Stavroulas, D.G.  Kaskaoutis, G.  Grivas, D.  Paraskevopoulou, U.C.  Dumka, M.  Tsagkaraki, A.  Bougiatioti, K.  Oikonomou, J. Sciare, E. Gerasopoulos, and N. Mihalopoulos, Long-term variability, source apportionment and spectral properties of black carbon at an urban background site in Athens, Greece, Atmos. Environ. 222, 117137 (2020), https://doi. org/10.1016/j.atmosenv.2019.117137

[15] V. Ulevicius, S. Byčenkienė, V. Remeikis, A. Garbaras, S.  Kecorius, J.  Andriejauskienė, D.  Jasinevičienė, and G.  Mocnik, Characterization of pollution events in the East Baltic region affected by regional biomass fire emissions, Atmos. Res. 98(2), 190–200 (2010), https://doi.org/10.1016/j.atmosres.2010.03.021

[16] S.  Byčenkienė, V.  Ulevicius, V.  Dudoitis, and J. Pauraitė, Identification and characterization of black carbon aerosol sources in the  East Baltic Region, Adv. Meteorol. 2013, 380614 (2013), https://doi.org/10.1155/2013/380614

[17] V. Ulevicius, S. Byčenkienė, C. Bozzetti, A. Vlachou, K.  Plauškaitė, G.  Mordas, V.  Dudoitis, G.  Abbaszade, V.  Remeikis, A.  Garbaras, et  al., Fossil and non-fossil source contributions to atmospheric carbonaceous aerosols during extreme spring grassland fires in Eastern Europe, Atmos. Chem. Phys. 16(9), 5513–5529 (2016), https://doi.org/10.5194/acp-16-5513-2016

[18] S.  Byčenkienė, V.  Ulevicius, and S.  Kecorius, Characteristics of black carbon aerosol mass concentration over the East Baltic region from twoyear measurements, J. Environ. Monit. 13(4), 1027–1038 (2011), https://doi.org/10.1039/C0EM00480D

[19] W.M.  Kirk, M.  Fuchs, Y.  Huangfu, N.  Lima, P.  O’Keeffe, B.  Lin, T.  Jobson, S.  Pressley, V. Walden, D. Cook, and B.K. Lamb, Indoor air quality and wildfire smoke impacts in the Pacific Northwest, Sci. Technol. Build. Environ. 24(2), 149–159 (2018), https://doi.org/10.1080/2374473 1.2017.1393256

[20] S. Brasche and W. Bischof, Daily time spent indoors in German homes  –  Baseline data for the  assessment of indoor exposure of German occupants, Int. J. Hyg. Environ. Health 208(4), 247–253 (2005), https://doi.org/10.1016/j.ijheh.2005.03.003

[21] B.F. Yu, Z.B. Hu, M. Liu, H.L. Yang, Q.X. Kong, and Y.H.  Liu, Review of research on air-conditioning systems and indoor air quality control for human health, Int. J. Refrig. 32(1), 3–20 (2009), https://doi.org/10.1016/j.ijrefrig.2008.05.004

[22] J.L.  Adgate, G.  Ramachandran, G.C.  Pratt, L.A.  Waller, and K.  Sexton, Spatial and temporal variability in outdoor, indoor, and personal PM2.5 exposure, Atmos. Environ. 36(20), 3255–3265 (2002), https://doi.org/10.1016/S1352-2310(02)00326-6

[23] L. Kliucininkas, E. Krugly, I. Stasiulaitiene, I. Radziuniene, T.  Prasauskas, A.  Jonusas, V.  Kauneliene, and D.  Martuzevicius, Indoor–outdoor levels of size segregated particulate matter and mono/polycyclic aromatic hydrocarbons among urban areas using solid fuels for heating, Atmos. Environ. 97, 83–93 (2014), https://doi.org/10.1016/j.atmosenv.2014.08.010

[24] J.  Zhang and K.R.  Smith, Indoor air pollution: a  global health concern, Br. Med. Bull. 68(1), 209–225 (2003), https://doi.org/10.1093/bmb/ldg029

[25] A.  Spinazzè, D.  Campagnolo, A.  Cattaneo, P.  Urso, I.A.  Sakellaris, D.E.  Saraga, C.  Mandin, N.  Canha, R.  Mabilia, E.  Perreca, et  al., Indoor gaseous air pollutants determinants in office buildings – The OFFICAIR project, Indoor Air 30(1), 76–87 (2020), https://doi.org/10.1111/ina.12609

[26] L. Du, V. Leivo, T. Prasauskas, M. Täubel, D. Martuzevicius, and U.  Haverinen-Shaughnessy, Effects of energy retrofits on Indoor Air Quality in multifamily buildings, Indoor Air 29(4), 686–697 (2019), https://doi.org/10.1111/ina.12555

[27] I. Stasiulaitiene, E. Krugly, T. Prasauskas, D. Ciuzas, L. Kliucininkas, V. Kauneliene, and D. Martuzevicius, Infiltration of outdoor combustiongenerated pollutants to indoors due to various ventilation regimes: A  case of a  single-family energy efficient building, Build. Environ. 157, 235–241 (2019), https://doi.org/10.1016/j.buildenv.2019.04.053

[28] B. Kozielska, A. Mainka, M. Żak, D. Kaleta, and W. Mucha, Indoor air quality in residential buildings in Upper Silesia, Poland, Build. Environ. 177, 106914 (2020), https://doi.org/10.1016/j.buildenv.2020.106914

[29] A. Garbaras, J. Šapolaitė, I. Garbarienė, Ž. Ežerinskis, A. Mašalaitė-Nalivaikė, R. Skipitytė, and A.  Plukis, Aerosol source (biomass, traffic and coal emission) apportionment in Lithuania using stable carbon and radiocarbon analysis, Isotopes Environ. Health Stud. 54(5), 1–12 (2018), https://doi.org/10.1080/10256016.2018.1509074

[30] M.  Collaud Coen, E.  Weingartner, A.  Apituley, D. Ceburnis, R. Fierz-Schmidhauser, H. Flentje, J.S.  Henzing, S.G.  Jennings, M.  Moerman, A.  Petzold, O.  Schmid, and U.  Baltensperger, Minimizing light absorption measurement artifacts of the  Aethalometer: evaluation of five correction algorithms, Atmos. Meas. Tech. 3(2), 457–474 (2010), https://doi.org/10.5194/amt-3-457-2010

[31] E. Weingartner, H. Saathoff, M. Schnaiter, N. Streit, B.  Bitnar, and U.  Baltensperger, Absorption of light by soot particles: determination of the absorption coefficient by means of aethalometers, J. Aerosol Sci. 34(10), 1445–1463 (2003), https://doi.org/10.1016/S0021-8502(03)00359-8

[32] J. Sandradewi, A.S.H. Prévôt, S. Szidat, N. Perron, M.R.  Alfarra, V.A.  Lanz, E.  Weingartner, and U. Baltensperger, Using aerosol light absorption measurements for the quantitative determination of wood burning and traffic emission contributions to particulate matter, Environ. Sci. Technol. 42(9), 3316–3323 (2008), https://doi.org/10.1021/es702253m

[33] P.  Zotter, H.  Herich, M.  Gysel, I.  El-Haddad, Y. Zhang, G. Močnik, C. Hüglin, U. Baltensperger, S.  Szidat, and A.S.H.  Prévôt, Evaluation of the  absorption Ångström exponents for traffic and wood burning in the  Aethalometer-based source apportionment using radiocarbon measurements of ambient aerosol, Atmos. Chem. Phys. 17(6), 4229–4249 (2017), https://doi.org/10.5194/acp-17-4229-2017

[34] O.  Favez, I.  El  Haddad, C.  Piot, A.  Boréave, E. Abidi, N. Marchand, J.L. Jaffrezo, J.L. Besombes, M.B.  Personnaz, J.  Sciare, H.  Wortham, C.  George, and B.  D’Anna, Inter-comparison of source apportionment models for the estimation of wood burning aerosols during wintertime in an Alpine city (Grenoble, France), Atmos. Chem. Phys. 10(12), 5295–5314 (2010), https://doi.org/10.5194/acp-10-5295-2010

[35] R.M.  Harrison, D.C.S.  Beddows, A.M.  Jones, A.  Calvo, C.  Alves, and C.  Pio, An evaluation of some issues regarding the  use of aethalometers to measure woodsmoke concentrations, Atmos. Environ. 80, 540–548 (2013), https://doi.org/10.1016/j.atmosenv.2013.08.026

[36] P. Paatero, A weighted non-negative least squares algorithm for three-way ‘PARAFAC’ factor analysis, Chemometr. Intell. Lab. Syst. 38(2), 223–242 (1997), https://doi.org/10.1016/S0169-7439(97)00031-2

[37] E.  Kostenidou, B.-H.  Lee, G.J.  Engelhart, J.R.  Pierce, and S.N.  Pandis, Mass spectra deconvolution of low, medium, and high volatility biogenic secondary organic aerosol, Environ. Sci. Technol. 43(13), 4884–4889 (2009), https://doi.org/10.1021/es803676g

[38] A.F. Stein, R.R. Draxler, G.D. Rolph, B.J.B. Stunder, M.D.  Cohen, and F.  Ngan, NOAA’s HYSPLIT atmospheric transport and dispersion modeling system, Bull. Am. Meteorol. Soc. 96(12), 2059–2077 (2016), https://doi.org/10.1175/BAMS-D-14-00110.1

[39] G.  Rolph, A.  Stein, and B.  Stunder, Real-time environmental applications and display sYstem: READY, Environ. Model. Software 95, 210–228 (2017), https://doi.org/10.1016/j.envsoft.2017.06.025

[40] C.  Fountoukis and A.  Nenes, ISORROPIA II: a computationally efficient thermodynamic equilibrium model for K+–Ca2+–Mg2+–NH4+–Na+–SO42−–NO3–Cl–H2O aerosols, Atmos. Chem. Phys. 7(17), 4639–4659 (2007), https://doi.org/10.5194/acp-7-4639-2007

[41] Y. Chen, H. Shen, and A.G. Russell, Current and future responses of aerosol pH and composition in the U.S. to declining SO2 emissions and increasing NH3 emissions, Environ. Sci. Technol. 53(16), 9646–9655 (2019), https://doi.org/10.1021/acs.est.9b02005

[42] J.  Ding, P.  Zhao, J.  Su, Q.  Dong, X.  Du, and Y. Zhang, Aerosol pH and its driving factors in Beijing, Atmos. Chem. Phys. 19(12), 7939–7954 (2019), https://doi.org/10.5194/acp-19-7939-2019

[43] H.  Guo, L.  Xu, A.  Bougiatioti, K.M.  Cerully, S.L.  Capps, J.R.  Hite  Jr, A.G.  Carlton, S.H.  Lee, M.H. Bergin, N.L. Ng, A. Nenes, and R.J. Weber, Fine-particle water and pH in the  southeastern United States, Atmos. Chem. Phys. 15(9), 5211–5228 (2015), https://doi.org/10.5194/acp-15-5211-2015

[44] A.  Bougiatioti, I.  Stavroulas, E.  Kostenidou, P. Zarmpas, C. Theodosi, G. Kouvarakis, F. Canonaco, A.S.H.  Prévôt, A.  Nenes, S.N.  Pandis, and N.  Mihalopoulos, Processing of biomassburning aerosol in the  eastern Mediterranean during summertime, Atmos. Chem. Phys. 14(9), 4793–4807 (2014), https://doi.org/10.5194/acp-14-4793-2014

[45] H. Timonen, S. Carbone, M. Aurela, K. Saarnio, S.  Saarikoski, N.L.  Ng, M.R.  Canagaratna, M. Kulmala, V.-M. Kerminen, D.R. Worsnop, and R.  Hillamo, Characteristics, sources and watersolubility of ambient submicron organic aerosol in springtime in Helsinki, Finland, J. Aerosol Sci. 56, 61–77 (2013), https://doi.org/10.1016/j.jaerosci.2012.06.005

[46] M.  Crippa, F.  Canonaco, V.A.  Lanz, M.  Äijälä, J.D.  Allan, S.  Carbone, G.  Capes, D.  Ceburnis, M.  Dall’Osto, D.A.  Day, et  al., Organic aerosol components derived from 25 AMS data sets across Europe using a  consistent ME-2 based source apportionment approach, Atmos. Chem. Phys. 14(12), 6159–6176 (2014), https://doi.org/10.5194/acp-14-6159-2014

[47] C.J.  Hennigan, A.P.  Sullivan, J.L.  Collett  Jr, and A.L. Robinson, Levoglucosan stability in biomass burning particles exposed to hydroxyl radicals, Geophys. Res. Lett. 37(9) (2010), https://doi.org/10.1029/2010GL043088

[48] S. Zhou, S. Collier, D.A. Jaffe, N.L. Briggs, J. Hee, A.J.  Sedlacek  III, L.  Kleinman, T.B.  Onasch, and Q. Zhang, Regional influence of wildfires on aerosol chemistry in the western US and insights into atmospheric ageing of biomass burning organic aerosol, Atmos. Chem. Phys. 17(3), 2477–2493 (2017), https://doi.org/10.5194/acp-17-2477-2017

[49] J. Pauraite, K. Plauškaitė, V. Dudoitis, and V. Ulevicius, Relationship between the optical properties and chemical composition of urban aerosol particles in Lithuania, Adv. Meteorol. 2018, 1–10 (2018), https://doi.org/10.1155/2018/8674173

[50] J.P.S.  Wong, M.  Tsagkaraki, I.  Tsiodra, N.  Mihalopoulos, K. Violaki, M. Kanakidou, J. Sciare, A.  Nenes, and R.J.  Weber, Atmospheric evolution of molecular-weight-separated brown carbon from biomass burning, Atmos. Chem. Phys. 19(11), 7319–7334 (2019), https://doi.org/10.5194/acp-19-7319-2019


J. Pauraitė a, I. Garbarienė a, A. Minderytė a, V. Dudoitis a, G. Mainelis b, L. Davulienė a, I. Uogintė a, K. Plauškaitė a, S. Byčenkienė a

a Fizinių ir technologijos mokslų centras, Vilnius, Lietuva

b Rutgerso Naujojo Džersio valstijos universitetas, Naujasis Džersis, JAV


Pavasarį deginama žolė tampa oro taršos šaltiniu, tačiau deginimo įtaka pastatų oro kokybei nėra iki galo ištirta. 1  µm skersmens kietųjų dalelių (KD1) masės koncentracija buvo matuojama biuro Vilniuje patalpoje ir pastato išorėje (lauke) atitinkamai balandžio 25–26 ir 26–29 dienomis. Šiomis dienomis dėl žolės deginimo buvo kilę gaisrai kaimyninėse šalyse (Baltarusijoje, Ukrainoje ir Rusijoje), taip pat Vilniaus miesto prieigose esančiose vietovėse. Su gaisrais siejami išmetimai į aplinką buvo tiriami naudojant atgalinių oro masių trajektorijų (HYSPLIT) ir gaisrų žemėlapio (MODIS) metodus. KD1 cheminė sudėtis buvo išmatuota aerozolio cheminės sudėties monitoriumi (ACSM) ir aetalometru. Organinės aerozolio dalelės fiksuotos tiek biure, tiek ir lauke (>70 % viso KD1). Biure oro filtravimo sistema sulaikė iki 55 % KD1 frakcijos masės koncentracijos. Nors KD1 masės koncentracijos biure buvo mažesnės, tačiau nustatytos žemesnio pH aerozolio dalelės galėjo turėti neigiamą poveikį žmogaus sveikatai. Anglies turinčių aerozolio dalelių šaltinių analizė parodė, kad vidutiniškai 56 % šių dalelių masės koncentracijos biure galimai susidarė dėl deginamos žolės.