aerosol direct radiative forcing: seasonal variability … · 2019-04-27 · bucurestii noi (44°28...

Romanian Journal of Physics 64, 808 (2019)

AEROSOL DIRECT RADIATIVE FORCING: SEASONAL VARIABILITY IN BUCHAREST AREA, ROMANIA

G. MANOLACHE1,2, S. STEFAN2, G. IORGA3,* 1 Technical University of Civil Engineering of Bucharest, Bd. Lacul Tei nr. 122–124, 020396, Bucharest, Romania

2 University of Bucharest, Faculty of Physics, Dept. of Matter Structure, Atmospheric and Earth Physics and Astrophysics, P.O. Box MG-11, Magurele, 077125 Bucharest, Romania

3 University of Bucharest, Faculty of Chemistry, Dept. of Physical Chemistry (Physics Group), Regina Elisabeta 4–12, 030018, Romania

*Corresponding author, E-mail: [email protected]

Received March 13, 2018

Abstract. The focus of present paper is upon aerosol variability in Bucharest metropolitan area and its relationships with aerosol direct effect/forcing (DRE/DRF). The aerosol (particulate matter less than 10 µm, PM10) mass concentrations were collected at three sampling sites (within a grid-scale of a regional climate model) covering three types of pollution (urban, suburban, and rural) in Bucharest area during 1st of June 2014 to 31st of May 2015. The aerosol optical properties were computed using the Optical Properties of Aerosol and Clouds (OPAC) software package, based on Mie scattering theory. We observed relatively high levels of PM10 with moderate to low seasonal cycle. DRF values indicate a cooling aerosol effect and show annual variations due to the combined effects of variability of aerosol optical properties over Bucharest region and of local geophysical variables. A section dedicated to comparison of aerosol DRF and its efficiency in Bucharest area and worldwide is included.

Key words: aerosol radiative forcing, forcing efficiency, Eastern Europe.

1. INTRODUCTION

Aerosols, either from natural or anthropogenic sources, represent important drivers of climate change [1]. Aerosols modify the radiation budget of atmosphere-earth climate interacting with the incoming solar radiation either directly by scattering and absorption [2–5], or indirectly by acting as cloud condensation nuclei thereby changing the cloud properties and influencing the precipitation process [6–11]. Moreover, the radiative heating by absorbing aerosols such as black carbon may affect the regional temperature profile [12], atmospheric circulation and water cycle [13–15]. Apart from the climate, aerosols are able to affect the human health, ecosystems, air pollution/air quality and visibility, at local and regional scales, through the various and complex processes aerosols are involved in [16–18]. The energy imbalance of the Earth-atmosphere system due to an imposed perturbation is the radiative forcing (RF), with its sub-components: direct, indirect and semi-direct RF expressing the aerosol direct, indirect and semi-direct effect. The aerosol

Article no. 808 G. Manolache, S. Stefan, G. Iorga 2

direct radiative forcing (DRF) at a given altitude is defined as the aerosol-induced change in the net radiative flux, namely the difference between net (downward minus upward) radiative flux for an atmosphere with and without aerosols. Considerable effort has been dedicated to estimate the aerosol direct effect worldwide, in cloud-free and cloudy conditions, over land and over ocean. Current estimate of total annual global mean radiative forcing of the aerosol–radiation interaction in [1] is –0.35 Wm–2 but this is still obtained with large uncertainty (–0.85 to +0.15 Wm–2). Central European aerosols have a much larger (negative) forcing than the global average values [19]. As an example, total direct forcing in Austria estimated at 70% relative humidity is equal to –2.90 Wm–2 (Vienna), –2.21 Wm–2 (Graz), –6.17 Wm–2 (Linz), +0.86 Wm–2 (Mt. Rax, 1644 m a.s.l.) and +1.24 Wm–2 (Mt. Sonnblick, 3106 m a.s.l.). The significant heating was caused by the fraction of black carbon in aerosol samples over a high reflective surface.

To express the impact of the local aerosol amount on the radiative forcing under various aerosol optical depths, the concept of the radiative forcing efficiency (RFE) is often used. RFE represents a measure of the radiative forcing due to the aerosol mass burden in the atmospheric layer within the particles are distributed that produce a unit variation in AOD (W m–2 AOD–1):

DRFRFEAOD

(1)

For desert dust events, in the Mediterranean area, in summer, Mallet et al. (2016) indicate RFE values [20] between about –60 Wm–2 and –85 Wm–2, depending on the intensity of the event (the more intense the event, the higher value RFE), while Markowitz et al. (2002) reports –34 Wm–2 (top of the atmosphere, TOA) and –94 Wm–2 (surface, SFC) in reference [21]. In typical conditions, estimations of RFE at Europe continental scale in summertime ranges from –29 Wm–2 to –85 Wm–2 when AERONET data were considered in model simulations, and from –13 –26 Wm–2 (TOA) to –36 –68 Wm–2 (SFC) when estimations were obtained from model-satellite integrations [22]. At local scale, the following RFE values of –24 Wm–2 (TOA) and –68 Wm–2 (SFC) were calculated for an urban site, Vienna, Austria, whereas for a coastal site, Almeira, Spain, RFE has smaller values –11 Wm–2 (TOA) and –57 Wm–2 (SFC) [23]. However, the literature values of aerosol radiative efficiency presents large differences worldwide.

Despite the intensive research during the past few decades, the understanding of the ambient aerosol effects remains a challenge and this is largely due to their variability and heterogeneity. Moreover, the estimates of the aerosol direct effect requires extensive observations on local and regional factors, such as the reflectivity of the ground surface, opacity of the atmosphere, cloud types and their frequency of occurrence, information that can be obtained from field campaigns in selected regions.

3 Seasonal aerosol forcing in Bucharest area Article no. 808

Because Bucharest represents the most polluted city in Romania [24], previous studies in its metropolitan area were focused on pollution issues, some optical properties or some micro–morphological aspects of [18, 25–30]. So far, there is no study that explored the aerosol direct effect at the regional scale of Bucharest in terms of radiative forcing. The aim of this study is to estimate the direct radiative forcing of the aerosols and its efficiency in Bucharest area using as input data the results of a yearlong intensive campaign, from 1st of June 2014 to 31st of May 2015 and to look for the seasonal variability of local DRE in Bucharest area, at a scale of about 50 km × 50 km. The results will give an overview of the variation, if this occurs, of radiative forcing in a relatively small area, which is about the size of a grid scale of a typical regional climate model. We also compare our model estimates of aerosol direct forcing and efficiency with DRF and RFE values reported at other European and worldwide sites.

This paper is organized as follows. The details of the performed measurements at all sites are presented in Section 2, together with data processing methods. The results of the analysis are discussed in Section 3. The conclusions end the paper.

2. EXPERIMENTAL SITES AND METEOROLOGY

2.1. SITE DESCRIPTIONS, MEASUREMENT METHODS AND DATA USED

Bucharest (approx. 44°26 N, 26°06 E) is located in open plain and presents in the last few decades a rapid expanding economic and industrial development. The major sources of anthropogenic pollution consist of emissions from power plants, high level of transportation, numerous small-scale industries and domestic activities in addition to seasonal agricultural activities, including biomass burning in its outskirts [18]. Sometimes dust coming from long-range distances can occur during strong Saharan storms [31]. Information about city, air pollution pattern and trends, including links to meteorology is detailed elsewhere [18]. PM10 (particulate matter with aerodynamic diameter below 10 µm) aerosol samples were collected at three sites (Fig. 1):

Bucurestii Noi (44°2859 N, 26°0226 E; 93 m a.m.s.l.) is located in a dense residential area in the north-west, characterized by intense traffic. Here construction activities related to the underground network contributed additionally to pollution during our sampling period, apart from other smaller construction activities.

Magurele (44°2056 N, 26°0201 E; 75 m a.m.s.l.) is a small city located about 10 km south-west from Bucharest downtown, categorized as suburban site by the National Environmental Protection Agency with respect to the air pollution.


Surlari site (44°3418 N, 26°1854 E; 76 m a.m.s.l.) is located in a rural area north-east of Bucharest at about 40 km from the downtown area and about 38 (52) km from Bucurestii Noi (Magurele) sites. The aerosol sampled at Surlari will give the regional levels in the area. The field station is accessible by car but is closed to the public all the time. The nearest village from the field station is at about 3 km.

Fig. 1 – (color online) Sampling sites in Bucharest Greater Area.

Three sets of PM10 measurements using similar instrumentation (single-stage low-volume air samplers, LVS Sven Leckel, flow rate 2.3 m3h–1) were used in the present work.

Table 1

Sampling sites, aerosol types and components

Site name (type) Aerosol type Aerosol components (mass mixing ratios, %) Bucurestii Noi (urban) urban water soluble (56.3); insoluble (35.8); soot (7.9) Magurele (suburban) continental average water soluble (58.3); insoluble (39.6); soot (2.1) Surlari (rural) continental clean water soluble (59.1); insoluble (40.9)

The samples were collected daily (24 hours, from morning to morning) from

1st of June 2014 to 31st of May 2015 on quartz fiber filters, and stored in laboratory in controlled atmosphere (humidity 50 5%, temperature 20° 1°C) for at least 48 hours before weighing. PM10 mass deposits were obtained by gravimetric method, the microbalance repeatability equals to 1 μg·m−3 for 24-h averages. At all sites, samples were collected at about 4 m above the ground. The PM10 aerosol types at


the three stations are presented in Table 1, together with their components and mass mixing ratios [32].

2.2. PM10 MEASUREMENTS OVER BUCHAREST

The seasonal and spatial heterogeneity of PM10 daily measurements is represented in Figure 2. We obtained PM10 mean levels in the range of 36.15 17.71 µg m–3 (autumn) to 72.86 9.18 µg m–3 (spring) for Bucurestii Noi site, from 28.95 5.22 µg m–3 (summer) to 42.26 10.35 µg m–3 (autumn) for Magurele and varied from 16.67 3.8 µg m–3 (summer) to 28.21 8.64 µg m–3 (winter) for Surlari. It appears a marked different variability between PM10 levels at the urban site Bucurestii Noi versus the suburban and rural sites during the period June 2014 to May 2015: lowest values are registered in summer at suburban and rural sites, in autumn at the urban site, whereas highest mass concentrations are recorded in different seasons. A comparison between sites indicates highest PM10 values at Bucurestii Noi, followed by Magurele and Surlari during summer, winter and spring. Exception appears in autumn when the highest PM10 mass concentrations were registered at Magurele.

Fig. 2 – Seasonal PM10 mass concentrations during the time period 1st of June 2014 – 31st of May 2015

(Summer = JJA, Autumn = SON, Winter = DJF, Spring = MAM).

2.3. METEOROLOGICAL CONDITIONS DURING SAMPLING PERIOD

Hourly surface meteorological observations were obtained from the automatic weather station of the Baneasa airport, which is closest to our sampling site Bucurestii Noi (Fig. 3). Observations in the data set from this station were compared to our three times a day measurements at Bucurestii Noi and they were found to be in very good agreement (below 5% absolute differences). Daily variations of surface level temperature during the campaign period indicate there was an overall increase in the surface air temperature from June 2014 reaching 25–26°C in August 2014. In autumn of 2014, the temperature amplitude Tmax–Tmin is found to be as high as 20°C and thereafter decreased to as low as –14°C in January of 2015. Generally, winter 2014–2015 was warmer than others [33]. In spring, Tmax–Tmin is found to vary between 8 and 18°C and temperature has a tendency to increase, as it was expected.


Fig. 3 – (color online) Daily means of observed temperature (T, °C), relative humidity (RH, %),

pressure (hPa), mixed layer depth (MXD, m) and local cloud cover fraction. Relative humidity had variations related to season with maximum values in

winter (90%) and dropped to below 50% in August 2014 and April 2015. Figure 3 also shows the cloud cover fraction (CCF) at each site and mixed layer depth (MXD) during the campaign period. Local daily cloud cover fractions were obtained from daily averages of hourly modeled CCF by the mesoscale numerical weather prediction model ALARO and provided by the National Meteorology Administration [34]. The daily values of MXD were computed using HYSPLIT model [35] and they show a seasonal behavior. In summer of 2014, the MXD average value is about 1000 m, in autumn of 2014 it decreased sharply to 200 m, this value being close to the average value for winter 2014–2015. In spring of 2015, due to the quickly changes of atmospheric air circulations, the variations of MXD are higher but having an increasing tendency (starting even from February 2015) from below 200 m up to over 1 km. We note many days in November 2014–February 2015 with high nebulosity, correlated to low values of MXD. Seasonal frequency distributions of wind direction (not shown here) indicate the dominance of airflows from north-eastern and western directions. Wind speeds reached only rarely higher values, most of them indicating gentle and moderate breeze (bellow 10 km/h). For wind speeds up to 20 km h–1 the frequency was about 10%, and for speeds between 20 and 30 km h–1 frequency was only about 4%.


3. METHOD OF ESTIMATION OF DIRECT RADIATIVE FORCING

In present study, the aerosol optical parameters were computed using the Optical Properties of Aerosol and Clouds (OPAC) software package [32] based on Mie scattering theory. The measured PM10 mass concentrations were converted into number concentrations using the mass mixing ratios indicated in Table 1. Aerosols were assumed to be internally mixed and we computed their optical properties as a function of wavelength for 25 wavelengths in the spectral range from 0.25 µm to 4 µm, using the volume averaged mixing rule and the input values of densities and the wavelength-dependent refractive indices taken from the OPAC code. We considered a uniform vertical distribution within the entire boundary layer and we used daily mixed depth (MXD) data, as Section 2 indicates.

Table 2

Mean seasonal input parameters for calculation of aerosol direct radiative forcing

Summer (JJA 2014)

Autumn (SON 2014)

Winter (DJF 2014–2015)

Spring (MAM 2015)

F0 (W m–2) 393 3.62 387 23.36 381 26.18 396 25.04 Ta 0.79 0.03 0.76 0.15 0.74 0.08 0.78 0.04 ß 0.19 0.10 0.23 0.10 0.33 0.20 0.23 0.10 Rs (site Bucurestii Noi) 0.14 0.17 0.20 0.17 Rs (site Surlari) 0.15 0.20 0.35 0.18 Rs (site Magurele) 0.16 0.20 0.30 0.19

In order to investigate the effect of water uptake by PM10 samples on

radiative forcing, all optical parameters were calculated using OPAC for dry conditions and for 50%, 70%, 80%, 90% relative humidity. Using the combined DRF model of Haywood and Shine [4] for internal mixtures of absorbing and non-absorbing aerosol components homogeneously mixed within the boundary layer and the model of Iorga et al. [19] by which the wavelength dependence of the incoming solar radiation is assumed to follow Planck’s equation for a blackbody with a temperature T = 6000 K, we obtained the estimation of the spectrally averaged radiative forcing DRF (in W m–2) due to scattering and absorption by the aerosol loading:

λ 4 m4

( ) ( , )DRF

σT

DRF E T

(2)

1)(

1)(R2)(R1)(AOD)()CCF1)((T)(F)(DRF s2s

2a0

(3)

where denotes Stefan-Boltzmann constant (5.67 × 10–8 W m–2 K–4), F0-the incoming radiation flux, Ta the fractional transmittance of the atmosphere (counting for molecular absorption and Rayleigh scattering), -single scattering albedo, -fraction


of radiation scattered upwards by the aerosol, AOD-aerosol optical depth, Rs-albedo of the underlying Earth's surface, -wavelength, CCF-cloud cover fraction.

Table 2 indicates the characteristic seasonal values of F0, Ta, and Rs resulted from the campaign measurements combined with multi-annual (1961–2000) climatology data [33]. The modeling method was applied both for cloud-free and for all-sky conditions for each site at all wavelengths and for all values of relative humidity used previously in computations of optical parameters.

4. RESULTS AND DISCUSSION

4.1. LOCAL DIRECT RADIATIVE FORCING IN CLOUD-FREE CONDITIONS

The aerosol DRF in Bucharest area shows a marked different seasonal evolution by site and Figure 4a,b indicates the maximum seasonal forcing for the two extreme values of the relative humidity: 0% and 90% at each site. In both cloud-free and all-sky conditions, aerosols have a distinct negative direct radiative effect, indicating they all mostly scatter than absorb radiation. For cloud-free situations, the forcing values vary from about –0.13 Wm–1 to about –0.90 W m–1 in dry conditions and from about –2.0 Wm–1 to about –8.0 Wm–1 when the relative humidity is 90%. The water-soluble substances dominate the radiative forcing at all sites, but the presence of absorbing species (soot) at Bucurestii Noi (7.9%) and Magurele (2.1%) reduces the cooling effect of the aerosols. This soot aerosol effect is expected, because soot gives a larger specific absorption coefficient and, as soot percentage gets higher in the total aerosol mass, it will create a higher positive contribution to total forcing. Even in situations with high humidity and precipitations, we suppose the soot aerosol in Bucurestii Noi area is not completely removed due to the presence of at least a persistent local source (vehicular exhaust for example) which replenishes the soot mass loading in the atmosphere rather quickly whatever the season. Seasonality of DRF by site was investigated taking into account the effect that variations of ground cover (cement/asphalt, grass, dark soil, ice or snow) play in the Bucharest area. As Table 2 indicates, different values of ground albedo (Rs) were used. Even in the higher reflective ground conditions in winter, we did not find the change of the sign of the forcing as for Central European aerosols [19]. DRF values in Bucharest area remained negative irrespective of season, showing the contribution of the geographical variables is quite important to the total aerosol effect, whatever the clouds were present (see Section 4.2) or not.

Analysis of DRF seasonality by site reveals that forcing does not reach its highest values during DJF, when we expect to have a higher PM10 burden confined in a lower atmospheric layer, but in spring for Bucurestii Noi and Surlari and in summer for Magurele. Whereas the maximum/minimum values of forcing for


Bucurestii Noi were obtained for highest (spring)/ lowest (autumn) levels of PM10, in the case of Magurele and Surlari, DRF is highest when PM10 mass concentrations are minimum (summer), respectively intermediate values (spring).

Fig. 4 – (color online) Seasonal variation of mean DRF (W m–2) of PM10 aerosols at the top of mixed

layer over Bucharest in cloud-free (a, b) and all-sky (c, d) conditions for RH = 0% and 90%; standard deviation bars are indicated.

Therefore, the role of wind on PM10 levels was further examined from polar

graphs (Fig. 5a-d), together with the ventilation coefficient (VC (m2s–1), the product between the maximum mixing layer height and the mean wind speed within mixing layer). VC indicates the capability of the atmosphere to dilute and disperse pollutants over an area (Fig. 5e); it shows the transport rate of pollutants in the mixing layer. A value of VC below 6000 m2s–1 with wind speeds less than 4 ms–1 is an indicative for a high air pollution potential [36].

Figure 5a-d indicates that, with the exception of summer, the aerosol burden relates to almost all directions of air flows, and this is associated with presence of about 57% (SON) – 45% (DJF) – 50% (MAM) anticyclone systems over the area [37]. In summer, the anticyclonic-type air circulation is higher (65%) and the PM10 levels at all sites are mostly associated with light eastern flows.


The temporal variation of ventilation coefficient was analyzed using the segmentation algorithm uDP (based on Dynamic Programming), the Scheffe test and the SEGMENTER software [38] in order to detect a significant change in the mean of the time series of VC. The detected change points are significant at 99% level.

Fig. 5 – (color online) Polar graphs of seasonal PM10 by site (a-d) and e) daily variation

of ventilation coefficient during period 1st of June 2014 – 31st of May 2015.


There were found 4 dominant change-points in the VC temporal variation, which determined the following corresponding 5 time periods: 01.06.2014–27.09.2014 (at point 119, sampling day); 30.09.2014–30.01.2015 (244); 31.01.2015–31.03.2015 (303), 01.04.2015–04.04.2015 (308); 05.04.2015–31.05.2015. On clear days when the air is well-mixed due to strong vertical convection, as in period covering summer and the September 2014, VC was about 7870 m2s–1. From end of September to end of March, ventilation in Bucharest area was the lowest (VC = 2422 m2s–1); small wind speeds and low mixing layer indicate the heaviest pollution potential during our study period. Somewhat correlated with this period, the meteorology data indicate that relative humidity was many times as high as 95% and many overcast days appeared. This suggests that aerosol particles might have been scavenged in fog and cloud droplet in larger extent, absorption was pronounced and dominated the scattering process as water on an insoluble core increases the radiation absorption capability of the core of the aerosol particle [39]. The highest ventilation coefficient/aerosol dilution and dispersion occurred only in a short period of 5 days at the beginning of April when VC = 19698 m2s–1 and the air flows came from western sector (260°–280°) with speeds between 2.78–4.44 ms–1; after this period, VC reduced by about 2.4 times to VC = 8179 m2s–1 for the rest of the studied period.

4.2. LOCAL DIRECT RADIATIVE FORCING IN ALL-SKY CONDITIONS

The estimated seasonal mean direct radiative forcing (DRF) of aerosols in all-sky conditions is presented in Figure 4c,d. As we observe, there are substantial differences between clear-sky and all-sky DRF for all values of relative humidity and for all sampling sites. In the presence of clouds, DRF estimates are lower showing aerosol direct effect reduces in magnitude, since clouds reflect solar radiation back to space and reduce the radiation available for interactions with aerosols. The DRF values vary from about –0.07 Wm–2 to about –0.45 Wm–2 in dry conditions and from about –0.5 Wm–2 to about –4.91 Wm–2 when the relative humidity is 90%.

The aerosol DRF is again negative, as it was found in cloud-free situation, but the presence of clouds determines a decrease of forcing magnitude by 61% (Bucurestii Noi), 72% (Magurele), 71% (Surlari), dry conditions, winter season, whereas in summer the decrease is only about 25% (Bucurestii Noi), 57% (Magurele), 40% (Surlari). The increase of humidity up to 90% does not change significantly the magnitude of decrease, differences being below 3% for all sites, with the exception of 5% for Surlari in winter when the local ground albedo was higher than at the rest of the sites. Apart from the variations in cloud cover fraction by site, the different DRF reduction percentages inter-sites is also related to the different aerosol optical properties in the three aerosol data sets (more absorbing aerosols


are present in Bucurestii Noi area than in Magurele and Surlari) and to the local seasonal variations in geophysical variables in Table 2.

4.3. DIRECT RADIATIVE FORCING EFFICIENCY

In the following, we discuss the aerosol direct radiative effect in terms of radiative forcing efficiency (RFE), in order to see the impact of the aerosol amount on the radiative forcing under various aerosol optical depths at local scale. The mean seasonal aerosol forcing efficiency is expressed by the slope of the DRF-AOD relationship at the wavelength of 500 nm, and it is shown in Table 3.

The seasonal forcing efficiency was found to vary locally between –6.18 and –31.42 Wm–2 AOD–1 at Bucurestii Noi, between about –15 and –31 Wm–2AOD–1 at Magurele, and between about –26 and –35 Wm–2AOD–1 for Surlari site. Seasonal RFE estimations revealed small differences from the RFE in cloud-free conditions when clouds were present (a few to some percentages). Aerosol radiative forcing efficiency is dependent on the ambient relative humidity, as water uptake by aerosols has effects on aerosol size, density and refractive index, and hence on extinction coefficient, AOD, single scattering albedo and scattering phase function. Reference [40] shows that the increase in relative humidity leads to decrease in forcing efficiency both at the surface and atmosphere, but large discrepancies were observed between measured and modeled (by incorporating measured aerosol properties) radiative impacts at the Earth’s surface over Bangalore in southern India. Forcing efficiency values estimated from radiative flux measurements may differ by as much as 20% between winter and summer [41].

Table 3

Mean seasonal estimations of radiative forcing efficiency (RFE), in cloud-free (CCF = 0) and all-sky (CCF # 0) conditions for RH = 0% and RH = 90%; nd-not determined

Site Bucurestii Noi Magurele Surlari CCF = 0

Season/RH RH = 0% RH = 90% RH = 0% RH = 90% RH = 0% RH = 90% JJA –7.62 –24.73 –20.11 –29.24 –29.38 –32.81 SON –6.18 –24.22 –18.36 –28.03 –27.79 –30.93 DJF –10.67 –31.42 –14.89 –29.88 –26.14 –26.56 MAM –6.65 –25.98 –21.00 –31.17 –30.97 –35.08

CCF # 0 JJA –6.38 –20.88 –19.32 –28.08 –29.11 –33.12 SON –3.92 –15.21 –9.72 –14.90 –11.20 –12.92 DJF –5.66 –16.68 –3.40 nd –10.60 nd MAM –5.86 –22.89 –10.62 –12.16 –15.50 –19.81

We observe that RFE decreases by 3–4 times for the highest polluted site

Bucurestii Noi, by 1.5–2 times for the medium polluted site Magurele and only by


1.1–1.3 times for the rural Surlari, when the relative humidity increases from 0 to 90%. Clouds’ presence determines an absolute decrease of RFE estimations by as much as units or tens of W m–2, depending on the relative humidity. For a specified relative humidity, the RFE seasonality is quite well pronounced and follows somehow the seasonality of the forcing at each site. As an example, at Surlari site, in cloud-free and dry conditions, RFE reaches its lowest value in winter (–26.14 Wm–2AOD–1) and the highest in spring (–30.97 Wm–2AOD–1), whereas at Bucurestii Noi we have –24.22 Wm–2AOD–1 in autumn and –31.42 Wm–2AOD–1 in winter. We also observe that differences between the lowest and highest values in cloud-free conditions could be significantly higher than those previously observed [41]: present model results indicate values by as much as 58–85% (dry) and 77–88% (RH = 90%). Between RFE values winter versus summer we obtained differences by as much as 70–80%.

4.4. COMPARISON BETWEEN DRF SI RFE FOR BUCHAREST AND WORLDWIDE AEROSOL

Estimations of DRF in urban areas in Europe are available from various modeling studies. Our results fit well within the Central European aerosol estimations [19]. In summer, radiative forcing at surface has –2.5 to –3 Wm–2 over most parts of Europe, –3.5 Wm–2 to –4 Wm–2 over the Mediterranean and the North Sea, whereas for winter, DRF varies from –2 Wm–2 over Mediterranean Sea to a positive forcing of +2 Wm–2 over the Eastern Europe due to black carbon presence [42]. A more recent study [43] shows forcing of urban and continental aerosol types showed a domain-average DRF with –16.3 Wm–2 during the summer period (8 days) and of –6.3 Wm–2 during the winter period (11 days) at surface. At TOA, an average value of –1.0 Wm–2 (summer) and of –2.7 Wm–2 (winter) is determined. Estimations of Su et al. (2013) of all-sky aerosol direct radiative effect for summer in 2004 at surface (–8.05 Wm–2) and (–3.95 Wm–2) TOA from MODIS data over Northern Hemisphere [44]. The presence of clouds reduces the surface cooling by about 30% (to –7.2 Wm–2).

Forcing efficiency in present study also lies within the range of RFE observed in Europe and compare well with the existing previous studies worldwide. For European aerosol, an efficiency value of –44.8 Wm–2AOD–1 resulted in summer and of –19.77 Wm–2AOD–1 for winter simulation period (values at surface, [43]). Somewhat higher RFE values (–65 to –85 Wm–2AOD–1) were reported for Saharan dust episodes for sites in southern Europe in summertime [20]. Annual mean European RFE values [45] for sulfate at surface are: –20.4 Wm–2AOD–1 (all sky) and –24.6 Wm–2AOD–1 (clear sky), for particulate organic matter: –32.8 Wm–2AOD–1 (all sky) and –37.9 Wm–2AOD–1 (clear sky), while for black carbon aerosol RFE: –190.8 Wm–2AOD–1 (all sky) and –207.8 Wm–2AOD–1 (clear sky). European aerosols


have a seasonal cycle with maximum DRF values in summer and minimum in winter [42].

On global scale, all-sky DRF estimations from satellite observations and GOCART simulations [44] present different seasonal cycles: model estimations reveal DRF values with minimum (–0.1 Wm–2) in autumn and maximum (–0.24 Wm–2) in winter, while satellite observations lead to minimum DRF in spring (–0.44 Wm–2) and maximum values in autumn (–0.6 Wm–2). Negative surface forcing, one order of magnitude higher than ours, was identified in Bangalore, India, with a maximum DRF in spring, in the pre-monsoon period, a lowest DRF in winter, when black carbon aerosol contributes to around 5%, and relatively constant values in rest of the year [40]. RFE values at the top of the atmosphere for a high altitude site [46] were found to be in the range of –3 to –20 Wm–2AOD–1 for both PM1 and PM10 aerosols with lower values for finer particles, but significantly higher than RFE values in more polluted sites in Indo-Gangetic plain.

The differences in estimations for both DRF and RFE in all studies compared here appear due to aerosol variability and heterogeneity that determine variability in aerosol microphysical and optical properties, due to local or regional input geographical variables but also due to differences in measurements and modelling methods.

5. CONCLUSIONS

The present study analyzed a yearlong (June 2014 to May 2015) measurements of near-surface particulate matter (PM10) in order to capture the magnitudes and behavior of radiative forcing and its efficiency over the urban area of Bucharest and surrounding area.

Three seasonal forcing structures were observed under same meteorological conditions.

Radiative forcing efficiency decreases as the site is more polluted and the relative humidity becomes lower.

Overall, the annual variations of direct radiative forcing arise from the combined effects of variability of optical properties of PM10 mass concentrations over Bucharest region and of local geophysical variables.

The presence of various fractions of clouds also changes the distribution of DRE. PM10 variability is influenced by the seasonally changed meteorological

conditions observed during the entire campaign period (moderate temperatures, wind speeds and high humidity, combined with dominant northeasterly and westerly air circulations).

The annual ventilation coefficient is 5903 m2s–1, which in context of moderate and low wind speeds, below 10–11 km h–1 (3 ms–1), together with an increased absolute frequency of the anticyclonic systems (with the exception of spring) indicates


a high pollution potential of Bucharest area with non-negligible implications for the DRE, even if the city is located in open plain.

The outcomes of this study confirm the potential of DRF variations within a grid point of a regional climate model when a large city/metropolitan area fits within the model scale.

Acknowledgements. We thank researchers from INOE2000 for help in measurement campaign at Magurele site, and to National Meteorology Administration for meteorology data at WMO_15420 and for the cloud cover fraction data at each measurement site from ALARO model. Figure 1 was created using Google Earth.

REFERENCES

1. IPCC: Climate Change 2013: The Physical Science Basis, Cambridge, UK and NY, USA. 2. R.A. McCormick and J. H. Ludwig, Science 156, 1358–1359 (1967). 3. C.F. Bohren and D.R. Huffman, Absorption and Scattering of Light by Small Particles. John

Wiley & Sons, New York (1983). 4. J. M. Haywood and K.P. Shine, Geophys. Res. Lett. 22, 603–606 (1995). 5. Y.A. Eltbaakh, M.H. Ruslan, M.A. Alghoul, K. Sopian, T.M. Razykov, Renew. Sust. Energ. Rev.

16, 4264–4276 (2012). 6. S. Twomey, J. Atmos. Sci. 34, 1149– 1152 (1977). 7. B. A. Albrecht, Science 245, 1227–1230 (1989). 8. R. Pincus and M.A. Baker, Nature 372, 250–252 (1994). 9. A.S. Ackerman, O.S. Toon, D.E. Stevens, A.J. Heymsfield, V. Ramanathan, E.J. Welton, Science,

288, 1042–1047 (2000). 10. G. Iorga and S. Stefan, Atmosfera 20, 247–269 (2007). 11. Y. Liu, Y., P.H. Daum, H. Guo, Y. Peng, Environ. Res. Lett. 3, 045021(2008). 12. J. Hansen, M. Sato, R. Ruedy, J. Geophys. Res. 102, 6831–6864 (1997). 13. C.A. Gueymard, C.N.S. Laulainen, J.K. Vaughan, F.E. Vignola, Proceedings of ASES Annual

Conference, June 16–21 (2000). 14. S. Menon, J. Hansen, L. Nazarenko, Y. Luo, Science 297, 2250–2253 (2002). 15. J. Kim, Y.Gu, K.-N. Liou, R.J. Park, C.-K. Song, Clim. Change 111, 859–877 (2012). 16. C.A. Pope, D.W. Dockery, J. Air Waste Manag. Assoc. 56, 709–742 (2006). 17. D.L. Yue, M. Hu, Z.J. Wu, S. Guo, M.T. Wen et al., Atmos. Chem. Phys. 10, 9431–9439 (2010). 18. G. Iorga, C. Balaceanu (Raicu), S. Stefan, Atmos. Poll. Res. 6, 824–834 (2015). 19. G. Iorga, R. Hitzenberger, A. Kasper-Giebl, H. Puxbaum, J. Geophys. Res. 112, D01204 (2007). 20. M. Mallet, F. Dulac, P. Formenti, P. Nabat, J. Sciare et al., Atmos. Chem. Phys. 16, 455–504

(2016). 21. K.M. Markowicz, P.J. Flatau, R.V. Ramana, P.J. Crutzen, V. Ramanathan, Geophys. Res. Lett.

29, 1968 (2002). 22. H. Yu, Y.J. Kaufman, M. Chin, M.G. Feingold, L.A. Remer et al., Atmos. Chem. Phys. 6, 613–666

(2006). 23. H. Horvath, L.A. Arboledas, F.J. Olmo, O. Jovanovic, M. Gangl et al., J. Geophys. Res. 107,

4386 (2002). 24. G. Iorga, Air pollution monitoring: A case study from Romania, Book Chapter (28 pg) in Air

Quality Air Quality – Measurement and Modeling, Prof. Philip Sallis (Ed.), InTech, DOI: 10.5772/64919 (2016).

25. C. Radu, D. Nicolae, A. Nemuc, A.C. Talianu, AAS08: 2nd Advanced Atmospheric Aerosol Symposium 16, 345–352 (2008).


26. L. Mihai, S. Stefan, J. Atmos. Oceanic Techn. 28, 1307–1316 (2011). 27. E.A. Olaru, Z.Y. Offer, F. Ruta, I. Udrea, Environ. Monit. Asses. 184, 4737–4745 (2012). 28. L. Makra, I. Ionel, Z. Csepe, I. Matyasovszky, N. Lontis, F. Popescu, Z. Sumeghy, Sci. Tot. Environ.

458, 36–46 (2013). 29. G. Grigoras, S. Stefan, C. Rada, C. Grigoras, Atmos. Poll. Res. 7, 567–576 (2016). 30. S. Voinea, G. Manolache, G. Iorga, S. Stefan, Rom. Rep. Phys. 70, 705 (2018). 31. M. Gothard, A. Nemuc, C. Radu, S. Dascalu, Rom. Rep. Phys. 66, 509–519 (2014). 32. M. Hess, P. Koepke, I. Schult, Bull. Am. Meteorol. Soc. 79, 831– 844 (1998). 33. I. Sandu, V.I. Pescaru, I. Poiana, A. Geicu, I. Candea, D. Tastea, Clima României, Acad. Publ. House,

Bucharest, Romania (2008). 34. D.C. Bostan, E.F. Manea, S. Stefan, Rom. Rep. Phys. 67, 1117–1127 (2015) 35. G. Rolph, A. Stein, B. Stunder, Environ. Modell. Softw. 95, 210–228 (2017). 36. U.S. Iyer, P.E. Raj, J. Earth Syst. Sci. 122, 537–549 (2013). 37. N. Barbu, C. Burada, S. Stefan, F. Georgescu, Acta Geophys. 64, 510–520 (2016). 38. A. Gedikli, H. Aksoy, N.E. Unal, J. Hydroinformatics, 12, 3, 318–328 (2010). 39. M.Z. Jacobson, Geophys. Res. Lett., 27, 217–220 (2000). 40. S.K. Satheesh, V. Vinoj, K. Krishna Moorthy, Atmos. Environ 44, 5295–5304 (2016). 41. M.Z. Jacobson, Nature 409, 695–697 (2001). 42. E. Marmer, B. Langmann, K. Hungershofer, T. Trautmann, J. Geophys. Res. 112, D23S16 (2007). 43. J. Meier, I. Tegen, B. Heinhold, R. Wolker, Geophys. Res. Lett. 39, L09802 (2012). 44. W. Su, N.G. Loeb, G.L. Schuster, M. Chin, F.G. Rose, J. Geophys. Res. 118, 655–669 (2013). 45. H. Yu, M. Chin, J. West, C.S. Atherton, N. Bellouin et al., J. Geophys. Res. 118, 700–720 (2013). 46. U.C. Dumka, D.G. Kaskaoutis, Atmos. Environ. 94, 96–105 (2014).

aerosol direct radiative forcing: seasonal variability … · 2019-04-27 · bucurestii noi (44°28...

Documents