Aerosol effects on the UV irradiance in Santiago de Chile

Aerosol effects on the UV irradiance in Santiago de Chile

Atmospheric Research 149 (2014) 282–291 Contents lists available at ScienceDirect Atmospheric Research journal homepage: www.elsevier.com/locate/atm...

2MB Sizes 2 Downloads 130 Views

Atmospheric Research 149 (2014) 282–291

Contents lists available at ScienceDirect

Atmospheric Research journal homepage: www.elsevier.com/locate/atmos

Aerosol effects on the UV irradiance in Santiago de Chile R.R. Cordero a,⁎, G. Seckmeyer b, A. Damiani a, J. Jorquera a, J. Carrasco c, R. Muñoz d, L. Da Silva e, F. Labbe e, D. Laroze f a

Universidad de Santiago de Chile, Ave Bernardo O'higgins 3363, Santiago, Chile Leibniz Universität Hannover, Herrenhäuser Str. 2, 30419 Hannover, Germany Direccion Meteorologica de Chile, Ave. Portales 3450, Santiago, Chile d Universidad de Chile, Av. Blanco Encalada 2002, Santiago, Chile e Universidad Técnica Federico Santa María, Ave. España 1680, Valparaíso, Chile f Instituto de Alta Investigación, Universidad de Tarapacá, Arica, Chile b c

a r t i c l e

i n f o

Article history: Received 6 April 2014 Received in revised form 22 June 2014 Accepted 2 July 2014 Available online 9 July 2014 Keywords: UV spectroradiometry UV irradiance Climatology

a b s t r a c t Santiago de Chile (33°27′ S–70°41′ W) is a mid-latitude city of 6 million inhabitants with a complicated surrounding topography. Aerosol extinction in Santiago is determined by the semi-arid local climate, the urban pollution, a regional subsidence thermal inversion layer, and the boundary-layer wind airflow. In this paper we report on spectral measurements of the surface irradiance (at 290–600 nm wavelength range) carried out during 2013 in the heart of the city by using a double monochromator-based spectroradiometer system. These measurements were used to assess the effect of local aerosols, paying particular attention to the ultraviolet (UV) range. We found that the aerosol optical depth (AOD) exhibited variations likely related to changes in the subsidence thermal inversion and in the boundary-layer winds. Although the AOD at 350 nm typically ranged from 0.2 to 0.3, peak values of about 0.7 were measured. The AOD diminished with the wavelength and typically ranged from 0.1 to 0.2 at 550 nm. Our AOD data were found to be consistent with measurements of the particulate matter (PM) mass concentration. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Increases in surface ultraviolet (UV) irradiance may lead to adverse effects on the biosphere including terrestrial and aquatic ecosystems as well as public health (Douglass et al., 2011). Therefore, an improved understanding of the UV climate has become of great interest. Previous studies have shown that satellite-derived UV estimates are biased at polluted sites (Tanskanen et al., 2006, 2007; Weihs et al., 2008; Ialongo et al., 2008; Kazadzis et al., 2009a, 2009b;

⁎ Corresponding author at: Universidad de Santiago de Chile, Ave Bernardo O'higgins 3363, Santiago, Chile. Tel.: +56 9 89018916; fax: +56 2 27181299. E-mail address: [email protected] (R.R. Cordero).

http://dx.doi.org/10.1016/j.atmosres.2014.07.002 0169-8095/© 2014 Elsevier B.V. All rights reserved.

Ialongo et al., 2009; Buchard et al., 2008; Douglass et al., 2011; Cabrera et al., 2012; Damiani et al., 2013), and also at sites affected by desert dust intrusions (Anton et al., 2012). In the particular case of Santiago de Chile, differences usually larger than 30% under cloudless conditions have been reported between ground-based measurements and estimates of the UV index (UVI) retrieved from the Total Ozone Mapping Spectrometer (TOMS) and from the Ozone Measurement Instrument (OMI) (see Cabrera et al., 2012; Damiani et al., 2013). The significant differences between ground-based measurements and satellite estimates found over heavily populated areas underline the importance of the tropospheric extinctions on local UV climatology and the need of ground-based measurements aimed at characterizing the optical properties of local aerosols.

R.R. Cordero et al. / Atmospheric Research 149 (2014) 282–291

using a sunphotometer of the AErosol RObotic NETwork (AERONET, see Holben et al., 1998). This instrument operated in Santiago only between August 2001 and October 2002 (Escribano et al., 2014). Satellite-derived estimates of the AOD have also been questioned in the case of Santiago. Estimates of the AOD retrieved from the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite instrument have shown consistency when compared with AERONET measurements elsewhere (Levy et al., 2010; Bréon et al., 2011; Hyer et al., 2011; Mei et al., 2012). However in the case of Santiago, a poor correlation was reported after comparing AERONET measurements over the period of 2001–02 (at 550 nm) with estimates of the AOD retrieved from MODIS (Oyanadel et al., 2006). Also a poor correlation was found when MODIS-derived estimates of the AOD have been compared with particulate matter (PM) mass concentrations measured over the period of 2000–10 in Santiago (see Escribano et al., 2014). This latter result was unexpected since at other locations (where atmospheric turbidity is driven by aerosols), a good correlation between MODIS-derived data and PM concentration has been reported (see Schaap et al., 2009; Boyouk et al., 2010; Tsai et al., 2011; Estellés et al., 2012). In this paper, we report on the first quality-controlled spectral measurements of the UV and visible irradiance in Santiago. These measurements were carried out during 2013 by using a double monochromator-based spectroradiometer meant to comply with the recommendations of the World Meteorological Organization (WMO) (Seckmeyer et al., 2001) and the NDACC specifications (Wuttke et al., 2006). The AOD was retrieved from our spectral measurements and used to weigh up the local tropospheric extinction. Ground-based measurements of the PM mass concentrations, as well as AOD estimates retrieved from different satellites were used for further comparisons.

Aerosol loading in Santiago is affected by the surrounding topography, the semi-arid local climate, the urban pollution, the regional subsidence thermal inversion layer, and the boundary-layer winds (see Perez and Salini, 2008; Oyanadel et al., 2006; Gramsch et al., 2000, 2006, 2009). The dispersion of local aerosols is constrained by the complicated surrounding topography (see Fig. 1) and the regional subsidence thermal inversion layer, which increases the near-ground lower-tropospheric stability over Santiago and determines the near-surface aerosol climatology (see Muñoz and Alcafuz, 2012). Seasonal changes in the subsidence thermal inversion layer over Santiago lead to variations in the depth of the aerosol layer; indeed, in fall and winter the depth of the aerosol layer is appreciably smaller with respect to spring and summer (see Muñoz and Undurraga, 2010). This is consistent with urban pollution measurements that show not only an increment in average surface concentrations in winter, but also the occurrence of multi-day episodes of high particulate matter concentrations associated with the strengthening of the subsidence thermal inversion (see Rutllant and Garreaud, 1995). Although the effect of aerosols on the UV climatology in Santiago is significant, no ground-based measurements of the aerosol optical properties in the UV range have been previously reported. Part of the problem arises from the fact that accurate measurements of the aerosol extinctions in the UV region require quality-controlled measurements of the UV spectra. However, because of the scarcity of spectroradiometer systems that comply with the required standards, the spectral UV monitoring stations in the southern hemisphere are currently underrepresented in the existing international networks (such as the Network for the Detection of Atmospheric Composition Change NDACC, see Wuttke et al., 2006). In the past, it has been assumed that aerosol extinctions in the UV region can be estimated by extrapolation from their effects in the visible region. However, even in the visible range, no reliable data regarding the optical properties of the aerosols in Santiago have been reported. Ground-based measurements of the aerosol optical depth (AOD) were carried in Santiago more than a decade ago by

a)

283

2. Irradiance measurements Ground-based spectral measurements of the solar irradiance were carried out at the campus of the Universidad de

b)

c)

10

33.1

3000 33.2 20

2500

Santiago

40

Latitude Degrees

30

Elevation m

Latitude Degrees

33.3

2000 1500 1000

33.4

USACH Ohiggins

33.5 33.6

Santiago 50

60

500

75 70 65 60 Longitude Degrees

0 72.0

33.7 33.8

71.5

71.0

Longitude Degrees

70.5

71.0

70.9

70.8

70.7

70.6

70.5

70.4

70.3

Longitude Degrees

Fig. 1. a) Santiago is surrounded by a complicated topography. b) Elevation along latitude 33°27′ S; the USACH Campus is located in Santiago at 33°27′ S–70°41′ W; the shaded area indicates the sea. c) Dotted line indicates the metropolitan area of Santiago; the locations of the USACH Campus, the “Parque O'higgins” station as well as the Cerrillos district are indicated in the plot.

284

R.R. Cordero et al. / Atmospheric Research 149 (2014) 282–291

3. Effect of aerosols on the surface UV

Santiago de Chile (USACH, located in the heart of the city, 33°27′ S–70°41′ W), by using the so-called USACH spectroradiometer. This instrument is based on a double monochromator Bentham® DMc150F-U, 150 mm focal length, and 1800 lines/mm gratings, fitted with a photomultiplier (PMT) as a detector. Although it is not a NDACC-certified instrument, the USACH spectroradiometer complies with the WMO recommendations (Seckmeyer et al., 2001) and the NDACC specifications (Wuttke et al., 2006). The USACH spectroradiometer sampled the irradiance every 1 nm (in the range of 290–600 nm). Sampling the radiance at each wavelength took about 1 s and therefore a spectral scan took about 5 min. Note that although the NDACC recommends a Full Width at Half Maximum (FWHM) b 1 nm, we have chosen to carry out our spectral measurements by using a slit function with a FWHM = 2 nm. This allowed us to carry out faster spectral scans by using a longer sampling wavelength step (1 nm). Although deconvolution can slightly improve the bandwidth (Wuttke et al., 2006), it was not applied to our measurements. Scans were carried at a 30 min interval. Quality control of our spectral measurements involved the absolute calibration and spectral alignment of the system at a regular base (every week). The calibration was achieved by using a baffled 100 W quartz halogen lamp. Wavelength misalignments were assessed by using a mercury lamp. The applied shift correction ensured a wavelength accuracy b0.05 nm. Based on the certificate of the lamp and the transfer of calibrations, we estimated the uncertainty involved in the absolute calibration to be up to 4% for UVA wavelengths and up to 10% for UVB wavelengths (see Cordero et al., 2008a,b, 2013). Note that the uncertainty of the UV irradiance computed from measurements carried out by NDACC-certified instruments has been estimated to be about 5% (see for details Seckmeyer et al., 2001). Aimed at the AOD assessment, measurements of the direct irradiance were facilitated by a robotic solar tracker INTRA® (with gears of 14:1, maximum speed of 8.5 deg. per second) and an input optics consisting of a 230 mm long collimator tube (with a baffle for avoiding stray light). The field-of-view (FOV) was approximately 2°. We also carried a limited number of spectral measurements of both the global and the diffuse irradiance (by using a shadow ring). In both cases an input optics with a 180 deg.-FOV flat diffuser was used.

b) 50

11h

13h

15h

70

Directa W

20 05.10.2013 06.10.2013 07.10.2013 11.10.2013

10

11h

13h

15h

Local Time UTC 3

m2

UV Irradiance

W

m2

UV Irradiance

30

09h

11h

13h

15h

17h

10h

19h

Global

50 40 30

05.10.2013 06.10.2013 07.10.2013 11.10.2013

20

0

09h

11h

13h

15h

17h

Local Time UTC 3

19h

14h

350

X X X X X X

300 05.10.2013 06.10.2013 07.10.2013 08.10.2013 09.10.2013 10.10.2013 11.10.2013

250 200

16h

__ Surface Measurements

10

17h

12h X OMI

400

60

40

0

c)

17h

Ozone Dobson Units

a)

Previous studies have shown that tropospheric extinctions play a role in the inter-hemispherical differences in peak UV values (McKenzie et al., 2006, 2008; Cordero et al., 2014). In the case of Santiago, we found that the aerosol load significantly affected peak UV values (occurring around noon), but the detected effects were different even when comparing consecutive days. Fig. 2 shows some measurements of the direct UV irradiance (plot 2a) and the global UV irradiance (plot 2b), carried out under cloudless conditions in October 2013 (dates are indicated in the plots). Clear-sky scenes were confirmed by using a whole-sky cloud monitoring system (based on a CCD camera and a fish-eye lens). In Fig. 2a and b, the UV irradiance was computed from our spectral measurements by integrating the UV spectra in the range of 290– 400 nm. It can be observed in Fig. 2a and b that the irradiance exhibited significant variations around noon on some days. Indeed, the direct UV irradiance in Fig. 2a shows changes of about 20% around noon on 05.10.2013 (day.month.year). Variations of about 12% can also be observed on the global UV irradiance corresponding to the same day (see Fig. 2b). Since the measurements shown in Fig. 2a and b were carried out under cloudless conditions, the significant changes around noon in the UV radiation on 05.10.2013 (also on 11.10.2013 and less significantly on 07.10.2013) can only be due to either jumps in the Total Ozone Column (TOC), or sudden variations in the tropospheric extinction. The possibility of unlikely changes in the ozone was tested by retrieving the TOC from our spectral UV measurements; we applied a method that implied comparing the ratio (between irradiances measured at different wavelengths) with a synthetic chart of this ratio computed for a variety of ozone values (Stamnes et al., 1991; Anton et al., 2014). Fig. 2c shows the TOC progression for 5–11.10.2013 computed from our ground-based measurements; OMI-derived estimates of the TOC are also shown (see crosses in Fig. 2c). The detected differences between satellite-derived estimates and our groundbased measurements of the TOC were within the range of ±3%. These figures agree with those shown by Anton et al. (2014) that also compared TOC values derived from global Bentham data and satellite estimates at Granada (Spain) (see Anton et al., 2014).

10h

12h

14h

16h

Local Time UTC 3

Fig. 2. Ground-based measurements over the period of 05–11.10.2013 (day.month.year). a) Direct UV irradiance (cloudless conditions only). b) Global UV irradiance (cloudless conditions only). c) Total Ozone Column (TOC).

R.R. Cordero et al. / Atmospheric Research 149 (2014) 282–291

4. Aerosol optical depth (AOD)

The insignificant fluctuations that the ozone exhibited through the day (see Fig. 2c) allowed us to conclude that the variations in the irradiance around noon on 05.10.2013 (also on 11.10.2013 and less significantly on 07.10.2013; see Fig. 2a and b) are likely due to sudden changes in the tropospheric extinction linked (as explained in the next section) with the local meteorology (in particular with the winds). Of course, changes in the tropospheric extinction do not occur around noon only. Actually, changes in the tropospheric extinction occurred all through the day often leading to “asymmetries” in surface UV irradiances. These asymmetries are apparent when comparing ground-based measurements carried out in the morning with measurements performed in the afternoon at the same solar zenith angle (SZA). Fig. 3a shows a scatter plot of the UV irradiances measured in the morning (see horizontal axis) and in the afternoon (see vertical axis) through the year, at the same SZA. Colors in the plot indicate the season. Fig. 3a shows that regardless of the season, UV irradiances in the morning tend to be lower than UV irradiances sampled in the afternoon at the same SZA. Since the measurements were carried out under cloudless conditions (and the effect of unlikely changes in the TOC is very small when integrating in the range of 290–400 nm), we conclude that the asymmetries in Fig. 3a are likely due to variations in the aerosols through the day. Fig. 3b depicts the ratio between the UV irradiances shown in plot 3a. Fig. 3b shows that the asymmetries (induced by changes in the aerosol extinction) depend on the SZA. Indeed, Fig. 3b confirms that, regardless of the season, tropospheric extinctions tend to be greater in the morning than in the afternoon. As explained in the next section, this is also due to the wind pattern in the area. Fig. 3c allows weighting up the effect of the aerosols on the UV irradiance in Santiago. It shows the aerosol transmittance, taken as the ratio between global UV irradiances measured under cloudless conditions and modeled assuming aerosol-free conditions. As radiative transfer model, we used UVSPEC (Mayer and Kylling, 2005). As shown in Fig. 3c, the aerosol transmittance ranged from 0.75 to 0.95 through the year, with an average of about 0.86.

Although as shown above, the effect of aerosols on the surface UV in Santiago is significant, no ground-based measurements of the aerosol optical properties in the UV range have been previously reported. Answering this challenge, the AOD was retrieved from our quality-controlled spectral measurements by applying a method based on the comparison of the measured spectral irradiance with UV spectra computed by using a radiative transfer model (see for details Cordero et al., 2009). The retrieved value of the AOD is that leading to the best match between the measured and the computed spectra. Again, we used UVSPEC. As an example, Fig. 4a shows spectral measurements of the direct irradiance (carried out in Santiago on 28.08.2013 and on 14.06.2013) as well as the corresponding spectra modeled by using the retrieved values of the AOD. Fig. 4b better exposes the agreement between the ground-based spectral measurements and the UVSPEC-computed spectra; it depicts the ratio between the spectra shown in plot 4a. Indeed, the slight differences shown in Fig. 4b are within the uncertainty bounds of both the measurements (Cordero et al., 2008a) and the UVSPEC-computed spectra (Cordero et al., 2007). Fig. 4c shows the wavelength-dependent values of the AOD (retrieved from the spectral measurements in plot 4a). As indicated in the plot, we characterized the spectral variations in the retrieved AOD values by using Angström's law: AOD = βλ − α, where λ is the wavelength in micrometers and α (related to the particle size) and β (related to the aerosol concentration) are referred to as Angström parameters. The AOD retrievals shown in Fig. 4c were chosen because they allow depicting the dispersions that the aerosols can exhibit in Santiago. Indeed, as shown in Fig. 4c, both the concentration and the size of the particles significantly change in Santiago. Fig. 5 shows the AOD values (at 350 nm) retrieved from our spectral measurements on different days through 2013. The measurements confirm that the AOD is generally greater in the morning than in the afternoon (see Fig. 5a), which leads to the “asymmetries” in the surface UV shown in Fig. 3. However, as shown in Fig. 5b the AOD can also show

b)

30

Global

20 30

40

50

60

UV Irradiance, Morning W m

2

70

1.0

0.8

0.6 10

Measurements

1.2

Global

20

30

40

SZA Degrees

50

60

Model No Aerosols

40

20

1.4

Summer Fall Spring

UV Irradiance

50

10 10

c) 1.4

Morning

60

Summer Fall Spring

Afternoon

70

UV Irradiance

UV Irradiance, Afternoon W m

2

a)

285

Summer Fall Spring

1.2

Global

1.0

0.8

0.6

30

40

50

60

70

80

SZA Degrees

Fig. 3. a) Scatter plot of the UV irradiances measured through the year (under cloudless conditions) in the morning and in the afternoon at the same solar zenith angle (SZA). b) Ratio between UV irradiances measured through the year (under cloudless conditions) in the morning and in the afternoon at the same solar zenith angle (SZA). c) Ratio between UV irradiances measured through the year (under cloudless conditions) and modeled assuming aerosol-free conditions. The UVSPEC model was used.

286

R.R. Cordero et al. / Atmospheric Research 149 (2014) 282–291

a)

b)

c)

Fig. 4. a) Ground-based spectral measurements of the direct irradiance (on 28.08.2013 and on 14.06.2013) and the corresponding modeled spectra. b) Ratio between ground-based spectral measurements of the direct irradiance (on 28.08.2013 and on 14.06.2013) and the corresponding modeled spectra. c) Aerosol Optical Depth (AOD) retrieved from ground-based spectral measurements (on 28.08.2013 and on 14.06.2013). Solid lines indicate the averages while the shaded area indicates the uncertainty.

Santiago ranged from 0.6 to 0.8 through the year, with an average of about 0.71 and no significant seasonal variations. The retrieved SSA values are consistent with urban tropospheric aerosols (see Krotkov et al., 1998). In spite of the SSA, the significant seasonal differences in the AOD (apparent when comparing Fig. 6a and b) were expected since fall and winter in Santiago are associated with a sharp reduction in the wind speed and a severe increment in the Particulate Matter (PM) mass concentrations. For example, Fig. 7a shows the results of measurements of both the wind speed and PM2.5 concentration at Parque O’Higgins (also in downtown Santiago, see Fig. 1c). These measurements over the period of 2012–2013 show the significant changes in the wind speed that likely led to the differences observed when comparing averages in Fig. 6a and b. Wind also significantly changes through the day. In winter, airflow in downtown Santiago is normally low; although hourly averages increase through the day, they are rarely greater than 2 m/s. In summer, hourly averages of the wind speed can be greater than 2 m/s several hours a day (8–12 h) peaking at about 2.8 m/s in the afternoon (about 17 h LT). Fig. 7b shows the progression through the day of the average of the winds computed for winter and fall, as well as for summer and spring. It can be observed that the intensity of the winds is normally low early in the morning but begins to increase at some moment as the morning progresses. Since

significant increments around noon or, as shown in Fig. 5c, it can remain roughly constant through the day. The increments in the AOD around noon detected on 05.10.2013 and 11.10.2013 (see Fig. 5b) are consistent with the variations in the UV irradiance around noon shown in Fig. 2a and b. Also, the few changes in the AOD detected on 06.10.2013 (see blue line in Fig. 5c) are consistent with the nearly symmetric irradiances measured that day (see the blue line in Fig. 2a). As explained below, the variations in the AOD values through the day shown in Fig. 5 are significantly affected by the winds. Fig. 6 shows our AOD measurements in different months at USACH Campus. It shows the average and the variability of the AOD values (at 350 nm and at 550 nm) retrieved from our measurements in summer and spring (see Fig. 6a), and in fall and winter (see Fig. 6b). Solid lines indicate the averages while the shaded areas indicate the variability (the latter was taken as the standard deviation of the retrievals). We further exploited our ground-based measurements by retrieving the single scattering albedo (SSA). We applied the method described by Bais et al. (2005) and Buchard et al. (2011), which is based on the comparison of the measured global irradiance with UV spectra computed by using a radiative transfer model. The retrieved values of the SSA are those leading to the best match between the measured and the computed spectra. We found that the SSA (at 350 nm) in

13h

15h

17h

0.8

λ 350 nm

0.6

14.06.2013 27.08.2013 07.10.2013 09.10.2013

0.4 0.2 0.0

11h

13h

15h

Local Time UTC 3

17h

b)

c) 11h

13h

15h

0.8

λ 350 nm

0.6

27.03.2013 05.10.2013 10.10.2013 11.10.2013

0.4 0.2 0.0

11h

13h

11h

17h

15h

Local Time UTC 3

17h

Aerosol Optical Depth AOD

11h

Aerosol Optical Depth AOD

Aerosol Optical Depth AOD

a)

13h

15h

0.8

17h λ 350 nm 25.03.2013 28.08.2013 06.10.2013

0.6 0.4 0.2 0.0

11h

13h

15h

17h

Local Time UTC 3

Fig. 5. AOD at 350 nm measured on different days through the year 2013. a) Peak AOD values occurred early in the morning. b) Peak AOD values occurred around noon. c) AOD values remain roughly flat through the day.

R.R. Cordero et al. / Atmospheric Research 149 (2014) 282–291

a)

287

b)

Fig. 6. AOD values (at 350 nm and at 550 nm) measured at USACH Campus: a) in summer and spring and b) in fall and winter. Solid lines indicate the averages while the shaded area indicates the observed variability.

these winds consistently blow from the southwest (from the Cerrillos district, see Fig. 1c, an industrial area of the city that early in the morning is more polluted than downtown Santiago; see Gramsch et al., 2006) peak AOD values normally occur at USACH Campus shortly after winds begin to blow. As an example of the influence of the winds on the peak AOD values, Fig. 7c shows the wind speed measured on different days in October 2013. On 05.10.2013 moderate winds from the southwest carrying polluted air began to blow around 12 h LT in downtown Santiago leading to an increment on the AOD (see Fig. 5b) and affecting in turn the UV irradiance around noon (see Fig. 2a). On 09.10.2013 winds of similar intensity began to blow 1 h earlier (around 11 h LT) affecting both the AOD (see Fig. 5a) and the UV irradiance (see Fig. 2a) at USACH Campus. Finally, on 06.10.2013, stronger winds also from the southwest cleaned both the Cerrillos district and downtown Santiago, early in the morning such that relatively low AOD values were measured through the day (see Fig. 5c) and no significant changes driven by the aerosols were detected in the UV that day (see Fig. 2a). Furthermore, the significant differences in the AOD variability (also apparent when comparing Fig. 6a and b) were expected since Santiago's basin is affected by multi-day episodes (typically in fall and winter) of high particulate matter concentrations associated with the strengthening of the subsidence thermal inversion (see Rutllant and Garreaud,

b) 5

100

5

80

4 3

4 3

60

2

40

1

20

Jan.12

July 12

Jan.13

July 13 Dec.13

Wind Speed m s

Monthly Average PM2.5 D Daily Average PM2 P .5

PM 2.5 μ g m

Wind Speed m s

Monthly Average Wind Daily a Average e Wind

11h

5. Particulate matter (PM) mass concentrations The Ministry of the Environment in Chile maintains an Ambient Air Monitoring Network in Santiago (http://sinca. mma.gob.cl/). This network consists of several well-maintained TEOM® instruments that enable direct measurements of the

13h

15h

17h

6

2013 Summer Spring Fall Winter

3 2 1 0

c)

Wind Speed m s

a)

1995; Muñoz and Undurraga, 2010; Muñoz and Alcafuz, 2012). We confirmed that on some winter days the AOD at 350 nm peaked at 0.7, which increased the dispersion of detected AOD values and in turn enhanced the variability. Regarding the visible range, averages of AOD values at 550 nm computed for fall and winter for Santiago (see Fig. 6b) are higher than the corresponding averages computed from AOD retrievals at 78 AERONET sites (see Thomas et al., 2008). However, regardless of the season, AOD values in Santiago are within the range of values observed at other polluted locations worldwide. Indeed, the AOD values at 550 nm measured at AERONET sites range from 0.06 to 0.7 with an average of 0.22 (see Thomas et al., 2008). Although changes in emission patterns have been reported (see Gallardo et al., 2012), we found that both the average and the variability of AOD values in the visible range computed for summer (see Fig. 6a) are similar to those measured a decade ago in Santiago by using an AERONET sunphotometer (most of AERONET measurements in Santiago were carried out in summer; see Escribano, 2012).

11h

13h

5

2013

4

06.10.2013 05.10.2013 09.10.2013

15h

17h

15h

17h

3 2 1

11h

13h

15h

Local Time UTC 3

17h

0

11h

13h

Local Time UTC 3

Fig. 7. Measurements at “Parque O'higgins”. a) Wind speed and PM2.5 concentration over the period of 2012–2013. Seasonal changes are apparent. b) Wind speed measured at different times through the day during 2013. Solid lines indicate the averages while the shaded area indicates the observed variability. c) Wind speed measured on different days in October 2013.

288

R.R. Cordero et al. / Atmospheric Research 149 (2014) 282–291

than in other seasons (see Muñoz and Undurraga, 2010), which leads to increments in the PM concentration but not necessarily affects the AOD. Nearby Parque O'higgins, a Vaisala® CL31 ceilometer (see Münkel et al., 2007 for technical details) operates since March 2007 monitoring the depth of the aerosol layer. This instrument renders profiles of range-corrected attenuated backscatter intensities (in units of 1/(105·srad·km), hereafter referred to as backscatter units, bu) with vertical resolution of 20 m. The backscatter intensities rendered by this instrument at low altitudes have been found to be correlated with the PM concentrations measured at Parque O'higgins (Muñoz and Alcafuz, 2012). Indeed, the ceilometer backscatter intensities are correlated with the aerosol concentration at different altitudes. Since in Santiago the vertical distribution of the aerosols appreciably changes through the year (see Muñoz and Undurraga, 2010), we chose to compare our AOD measurements with the column of backscatter intensities (rather than with the backscatter intensity corresponding to a specific altitude). The Backscatter Intensity Column (BIC) was computed by summing up the intensities measured at altitudes that ranged from 40 m to 400 m. Fig. 10 shows the variations through the day (during several days in different seasons) of the BIC (see dotted lines) and the AOD measured at 350 nm (see bold lines). It can be observed in this figure that changes through the day in the AOD are associated with similar changes in the BIC. Moreover, note that in spite of what we found when comparing AOD and PM concentration (see Fig. 9b and c), the ratio between the AOD and the BIC did not significantly change through the year.

PM mass concentration at different sites in the metropolitan area of the city (including Parque O'higgins and Cerrillos). Details on the network have been published elsewhere (see for details Jorquera, 2002). The measurements of PM2.5 mass concentration at Parque O'higgins were used to test the consistency of our AOD measurements. Fig. 8 summarizes the measurements of PM2.5 concentration though 2013 at Parque O'higgins. It shows the average and the variability of the PM2.5 concentration computed for summer and spring (see Fig. 8a), and for fall and winter (see Fig. 8b). Again, solid lines indicate the averages while the shaded areas indicate the variability (the latter was taken as the standard deviation of the retrievals). Comparing Fig. 6 with Fig. 8, it can be observed that changes through the day in the PM2.5 concentration (see Fig. 8) seem to be consistent with the behavior detected in our measurements of the AOD (see Fig. 6). Moreover, the variability of the PM2.5 mass concentration (see Fig. 8) and the variability of our AOD measurements (see Fig. 6) are of the same order. Fig. 9a shows the scatter plot of PM2.5 concentrations and our AOD measurements (at 3 different wavelengths: 350 nm, 450 nm, 550 nm) through the year. Color indicates the wavelength. Dotted lines indicate the linear regression. A good correlation was found between our ground-based AOD measurements and the PM concentrations regardless of the compared wavelength. Fig. 9b and c shows variations through the day during several days in spring (see Fig. 9b) and in fall and winter (see Fig. 9c) of both the PM2.5 concentrations (dotted lines) and our AOD measurements at 350 nm (bold lines). It can be observed in these figures that changes through the day in the AOD are associated with similar changes in the PM2.5 concentration. However, note that the ratio between the AOD and the PM concentration changes through the year exhibiting seasonal variations. Indeed, in spring (see Fig. 9b) AOD values of about 0.3 are associated with PM concentrations of about 20 μg/m3; while in winter (see Fig. 9c) AOD values of about 0.3 are associated with PM concentrations two times higher (about 40 μg/m3). These seasonal differences were also expected since the PM concentration depends not only on the aerosol loading but also on the depth of the aerosol layer. As discussed above, the vertical distribution of the aerosols undergoes seasonal changes due to the strengthening of the thermal inversion. In fall and winter the depth of the aerosol layer is appreciably smaller

a) 120

11h

13h

2013 Summer

PM2.5

3

60 40 20

11h

13h

15h

17h

2013 Fall Winter

100

Spring

80

0

b)

PM2.5 μg m

3

17h

AOD estimates in the visible range retrieved from the Moderate Resolution Imaging Spectroradiometer (MODIS) (aboard Aqua satellite), from MODIS and Multi-angle Imaging SpectroRadiometer (MISR) (both aboard Terra satellite), and Ozone Measurement Instrument (OMI) (aboard Aura satellite), have been used for further comparisons. We used Level-3 MODIS (Hubanks et al., 2008) Collection 5.1 product with Dark-Target algorithm (Levy et al., 2007), OMAERUV version 3 daily level 3 product from OMI (Torres et al., 2007), and MIL3DAE daily level 3 product from MISR (Diner et al., 2001). The overpass time is in the morning for Terra satellite

120

100 gm

15h

6. Satellite-derived estimates of the AOD

80 60 40 20

11h

13h

15h

Local Time UTC 3

17h

0

11h

13h

15h

17h

Local Time UTC 3

Fig. 8. PM2.5 concentration measured at “Parque O'higgins”: a) in summer and spring and b) in fall and in winter. Solid lines indicate the averages while the shaded areas indicate the observed variability.

R.R. Cordero et al. / Atmospheric Research 149 (2014) 282–291

b)

120

2013 _ _ _ PM2.5

0.60

60 40

R2 0.64 R2 0.63

20

05.10.2013 06.10.2013

30 20

0.45 0.30

10

0.15

R2 0.64

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 11h

13h

15h

0.00 17h

13.06.2013 28.08.2013

80

0.75 0.60

60

0.45

40

0.30

20

0.15

0 11h

Local Time UTC 3

Aerosol Optical Depth AOD

____ AOD _ _ _ PM2.5

100 3

____ AOD

40

PM 2.5 μ g m

350 nm 450 nm 550 nm

Aerosol Optical Depth AOD

PM 2.5 μ g m

80

c)

3

3

100

PM 2.5 μ g m

120

13h

15h

Aerosol Optical Depth AOD

a)

289

0.00 17h

Local Time UTC 3

Fig. 9. a) Scatter plot of the PM2.5 concentrations and AOD measurements through the year. b) PM2.5 concentrations (see dotted lines) and AOD values at 350 nm (see bold lines) measured through the day (during two days in spring). c) PM2.5 concentrations (see dotted lines) and AOD values at 350 nm (see bold lines) measured through the day (during a day in fall and a day in winter).

0.8 ____ AOD _ _ _ Ceilometer 25.03.2013 05.10.2013 27.08.2013

6

0.3

4

0.4

2

0.2

0

11h

13h

15h

Aerosol Optical Depth AOD

8

3

Total Backscatter intensities x 10 bu

(10:30 h equator crossing time) while in the case of both Aqua and Aura satellites it is in the early afternoon (13:30 h and 13:45 h equator crossing time, respectively). Fig. 11a shows the monthly climatology of the satellitederived AOD at 550 nm for the region around Santiago computed over the period of 2005–2012. We used gridded satellite products with a spatial resolution of 1 × 1° lat/lon (except for MIRS data whose resolution is 0.5 × 0.5 deg). The monthly MODIS-based climatology shown in Fig. 11a is similar to that shown elsewhere (see Escribano et al., 2014) and built up by using Level-2 MODIS data with a better spatial resolution. Nevertheless, note that the Level-3 MODIS data change only slightly with the spatial resolution (Ruiz-Arias et al., 2013). The differences in satellite estimates shown in Fig. 11a can be related to several issues. For example, the differences between the AOD data from Terra and the retrievals from Aqua and Aura are expected, since these satellites have different morning/afternoon overpass times and the aerosol load in Santiago exhibits significant variations through the day. On the other hand, the differences between MODIS/Terra and MODIS/Aqua data in Fig. 11a could be at least partially explained by the well-known degradation of the optic sensor

17h

Local Time UTC 3 Fig. 10. Backscatter Intensity Column (see dotted lines) and AOD at 350 nm (see bold lines) measured through the day (during days in different seasons).

on Terra (Levy et al., 2010). Moreover, the different spatial resolution of MISR/Terra could also play a (minor) role. Fig. 11a shows that even if OMI follows the general annual trend of MODIS, it renders higher AOD values. In addition to a likely calibration offset with respect to MODIS data (Ahn et al., 2008), one of the main issues affecting the quality of OMI aerosol products is the sub-pixel cloud contamination caused by the relatively large footprint of the OMI observations (i.e. 13 × 24 km2 at nadir). Therefore, clouds could contribute to enhance OMI AOD values. Since the differences between OMI and MODIS are smallest during summer months (characterized by clear sky conditions; see Damiani et al., 2014), it is likely that the cloud effect increases OMI-derived AOD values in winter/fall. The four series of satellite-based AOD data exhibit in general terms a similar behavior: i.e. lower AOD values in winter/fall months and higher values in summer and spring. Although MISR data show smaller winter/summer differences than other datasets, this could be related to the different spatial resolutions (i.e. 0.5 × 0.5 vs. 1 × 1° lat/lon). This behavior is different from that observed in our AOD measurements that in turn agree with long term PM2.5 measurements (see Fig. 11b): higher AOD values in fall and in winter, while lower AOD values in spring and in summer. Our AOD measurements though 2013 also suggest that satellite-derived estimates of the AOD may be biased low for fall/winter months, while appear to be somehow biased high for spring/summer months. These findings reinforce previous efforts that have shown poor correlations when comparing MODIS-derived estimates of the AOD with AERONET measurements over the period of 2001–02 (at 550 nm) (Oyanadel et al., 2006) and with PM mass concentrations measured over the period of 2000–10 in Santiago (Escribano et al., 2014). Although incongruences between ground-based and satellite data may be linked with the surface reflectance of the area (Levy et al., 2010; Escribano et al., 2014), additional long-term ground-based observations are needed to confirm this hypothesis. 7. Summary and conclusions Ground-based spectral measurements of the solar irradiance (at 290–600 nm wavelength range) were carried out in

290

R.R. Cordero et al. / Atmospheric Research 149 (2014) 282–291

a)

b) 2005 2012

80

λ 550 nm MODIS Aqua MODIS Terra OMI Aura MISR Terra

0.4 0.3 0.2

2005 2012 3

0.5

PM 2.5 μg m

Aerosol Optical Depth AOD

0.6

60

40

20

0.1 1

2

3

4

5

6

7

8

9 10 11 12

0

1 2 3 4 5 6 7 8 9 10 11 12

Month

Month

Fig. 11. a) Monthly climatology of the AOD for the region around Santiago computed over the period of 2005–2012 by using four different satellite-derived data. b) Monthly climatology of the PM2.5 concentration measured in downtown Santiago (at Parque O'higgins) over the period of 2005–2012.

the heart of Santiago, a mid-latitude city of 6 million inhabitants, whose aerosol loading is determined by the semi-arid local climate, the urban pollution, the regional subsidence thermal inversion layer, and the boundary-layer wind airflow. Our measurements (carried out during 2013) were meant to track the effect of aerosols on the surface irradiance paying particular attention to the UV range. We found that the tropospheric extinction (i.e. the AOD) tends to be greater in winter than in summer. These seasonal changes seem to be related to both the strengthening of the subsidence thermal inversion and the diminishment in wind speeds, which occur in fall and in winter. In winter, airflow in downtown Santiago is normally low and is rarely greater than 2 m/s. Instead, in summer, hourly wind speed averages can be greater than 2 m/s for several hours a day (8–12 h). Significant seasonal differences in the AOD variability (that was considerably greater in winter than in summer) were detected. The increment in AOD variability in winter was expected since Santiago's basin is affected by multi-day episodes (typically in fall and winter) of high particulate matter concentrations associated with the strengthening of the subsidence thermal inversion. We confirmed that on some winter days the AOD at 350 nm peaked at 0.7, which increased the dispersion of detected AOD values and in turn enhanced the variability. We also found significant variations through the day in the AOD values (that were generally greater in the morning than in the afternoon). These changes seem to be also driven by the winds. The intensity of the airflow in downtown Santiago is normally low early in the morning but begin to increase at some moment as the morning progresses. Since these winds consistently blow from the southwest (from an industrial area of the city that early in the morning is more polluted) peak AOD values normally occur in downtown Santiago shortly after winds begin to speed up. We found that variations in the wind speed around noon (and the resulting changes in the AOD easily detectable in spring and summer) significantly affect peak UV values. For example, we detected sudden increments in the AOD in the UV range (from 0.25 to 0.45) that led to reductions of about 20% in the direct UV irradiance (about 12% in the global UV irradiance) around noon. Changes in the AOD with the wavelength were also detected; while at 350 nm the AOD typically ranged from 0.2 to 0.3, and at 550 nm the AOD ranged from 0.1 to 0.2. Our ground-based measurements of the AOD were found to be consistent with measurements of the particulate matter (PM) mass concentration.

Acknowledgments The support of CONICYT-REDES (Preis 130047), CONICYTBMBF (Preis PCCI20130041), FONDEF (Preis IT13I10034), FONDECYT (Preis 1120639, Preis 1140239 and Preis 1120764), Millennium Scientific Initiative (Preis P10-061-F), CEDENNA, UTA-Project 8750-12, USACH-DICYT ASOCIATIVO, and UTFSMDGIP, is gratefully acknowledged. The measurements of particulate matter (PM) mass concentration were provided by Mr. Marcelo Corral from the Ministerio del Medio Ambiente (Chile).

References Ahn, C., Torres, O., Bhartia, P.K., 2008. Comparison of ozone monitoring instrument UV aerosol products with Aqua/moderate Resolution Imaging Spectroradiometer and Multiangle Imaging Spectroradiometer observations in 2006. J. Geophys. Res. 113. http://dx.doi.org/10.1029/ 2007JD008832 (D16S27). Anton, M., Valenzuela, A., Roman, R., Lyamani, H., Krotkov, N., Arola, A., Olmo, F. J., Alados-Arboledas, L., 2012. Influence of desert dust intrusions on ground-based and satellite-derived ultraviolet irradiance in southeastern Spain. J. Geophys. Res. 117. http://dx.doi.org/10.1029/2012JD018056 (D19209). Anton, M., Mateos, D., Roman, R., Valenzuela, A., Alados-Arboledas, L., Olmo, F.J., 2014. A method to determine the ozone radiative forcing in the ultraviolet range from experimental data. J. Geophys. Res. Atmos. 119, 1860–1873. http://dx.doi.org/10.1002/2013JD020444. Bais, A.F., Kazantzidis, A., Kazadzis, S., Balis, D.S., Zerefos, C.S., Meleti, C., 2005. Deriving an effective aerosol single scattering albedo from spectral surface UV irradiance measurements. Atmos. Environ. 39 (6), 1093–1102. Boyouk, N., Leon, J., Delbarre, H., Podvin, T., Deroo, C., 2010. Impact of the mixing boundary layer on the relationship between PM2.5 and aerosol optical thickness. Atmos. Environ. 44, 271–277. Bréon, F.M., Vermeulen, A., Descloitres, J., 2011. An evaluation of satellite aerosol products against sunphotometer measurements. Remote Sens. Environ. 115, 3102–3111. Buchard, V., Brogniez, C., Auriol, F., Bonnel, B., Lenoble, J., Tanskanen, A., Bojkov, B., Veefkind, P., 2008. Comparison of OMI ozone and UV irradiance data with ground-based measurements at two French sites. Atmos. Chem. Phys. 8, 4517–4528. Buchard, V., Brogniez, C., Auriol, F., Bonnel, B., 2011. Aerosol single scattering albedo retrieved from ground-based measurements in the UV and visible region. Atmos. Meas. Tech. 4, 1–7. Cabrera, S., Ipina, A., Damiani, A., Cordero, R., Piacentini, R., 2012. UV index values and trends in Santiago, Chile based on ground and satellite data. J. Photochem. Photobiol. B 115, 73–84 C. Cordero, R., Seckmeyer, G., Pissulla, D., DaSilva, L., Labbe, F., 2007. Uncertainty evaluation of the spectral UV irradiance evaluated by using the UVSPEC radiative transfer model. Opt. Commun. 276, 44–53. Cordero, R., Seckmeyer, G., Pissulla, D., DaSilva, L., Labbe, F., 2008a. Uncertainty evaluation of spectral UV irradiance measurements. Meas. Sci. Technol. 19, 1–15.

R.R. Cordero et al. / Atmospheric Research 149 (2014) 282–291 Cordero, R., Seckmeyer, G., Pissulla, D., Labbe, F., 2008b. Uncertainty of experimental integrals: application to the UV index calculation. Metrologia 45, 1–10. Cordero, R., Seckmeyer, G., Pissulla, D., Labbe, F., 2009. Exploitation of spectral direct UV irradiance measurements. Metrologia 46, 19–25. Cordero, R., Seckmeyer, G., Damiani, A., Labbe, F., Laroze, D., 2013. Monte Carlo-based uncertainties of surface UV estimates from models and from spectroradiometers. Metrologia 50, L1–L5. Cordero, R., Seckmeyer, G., Riechelmann, S., Damiani, A., Labbe, F., Laroze, D., 2014. The world's highest levels of surface UV: a case study. Photochem. Photobiol. Sci. 13, 70–81. Damiani, A., Cabrera, S., Muñoz, R., Cordero, R., Labbe, F., 2013. Satellitederived UV irradiance for a region with complex morphology and meteorology: comparison against ground measurements in Santiago de Chile. Int. J. Remote Sens. 34, 5812–5833. Damiani, A., Cordero, R., Cabrera, S., Laurenza, M., Rafanelli, R., 2014. Cloud cover and UV index estimates in Chile from satellite-derived and ground-based data. Atmos. Res. 138, 139–151. http://dx.doi.org/10. 1016/j.atmosres.2013.11.006. Diner, D.J., et al., 2001. MISR aerosol optical depth retrievals over southern Africa during the SAFARI-2000 dry season campaign. Geophys. Res. Lett. 28, 3127–3130. Douglass, A., Fioletov, V., Godin-Beekmann, S., Müller, S., Stolarski, R., Webb, A., Arola, A., 2011. Stratospheric ozone and surface ultraviolet radiation 1–80. Global Ozone Research and Monitoring Project-Report No. 52 (516 pp., Geneva, Switzerland). Escribano, J., 2012. Capa Limite, Reflectancia y Espesor Optico de Aerosoles Sobre Santiago(In Spanish). Master Thesis Universidad de Chile, Santiago, Chile. Escribano, J., Gallardo, L., Rondanelli, R., Choi, Y., 2014. Satellite retrievals of aerosol optical depth over a subtropical urban area: the role of stratification and surface reflectance. Aerosol Air Qual. Res. 14 (3), 596U568. Estellés, V., Martínez-Lozano, J., Pey, J., Sicard, M., Querol, X., Esteve, A., Utrillas, M., Sorribas, M., Gangoiti, G., Alastuey, A., Rocadesnbosch, F., 2012. Study of the correlation between columnar aerosol burden. Suspended matter at ground and chemical components in a background European environment. J. Geophys. Res. 117 (D04201). Gallardo, L., Escribano, J., Dawidowski, L., Rojas, N., Andrade, M., Osses, M., 2012. Evaluation of vehicle emission inventories for carbon monoxide and nitrogen oxides for Bogotá, Buenos Aires, Santiago and São Paulo. Atmos. Environ. 47, 12–19. Gramsch, E., Catalán, L., Ormeño, I., Palma, G., 2000. Traffic and seasonal dependence of the light absorption coefficient in Santiago de Chile. Appl. Opt. 39 (27), 4895–4901. Gramsch, E., Cereceda-Balic, F., Oyola, P., Baer, D., 2006. Examination of pollution trends in Santiago de Chile with cluster analysis of PM10 and ozone data. Atmos. Environ. 40, 5464–5475. Gramsch, E., Gidhagen, L., Wahlin, P., Oyola, P., Moreno, F., 2009. Predominance of soot-mode ultrafine particles in Santiago de Chile: possible sources. Atmos. Environ. 43, 2260–2267. Holben, B.N., et al., 1998. AERONET—a federated instrument network and data archive for aerosol characterization. Remote Sens. Environ. 66, 1–16. Hubanks, P., King, M., Platnick, S., Pincus, R., 2008. MODIS atmosphere L3 gridded product algorithm theoretical basis document Collection 005 Version 1.1. Tech. Rep. ATBD-MOD-30. NASA. Hyer, E., Reid, J., Zhang, J., 2011. An over-land aerosol optical depth data set for data assimilation by filtering, correction and aggregation of MODIS Collection 5 optical depth retrievals. Atmos. Meas. Tech. 4, 379–408. Ialongo, I., Casale, G.R., Siani, A., 2008. Comparison of total ozone and erythemal UV data from OMI with ground-based measurements at Rome station. Atmos. Chem. Phys. 8, 3283–3289. http://dx.doi.org/10.5194/acp-8-3283-2008. Ialongo, I., Buchard, V., Brogniez, C., Casale, G.R., Siani, A., 2009. Aerosol Single Scattering Albedo retrieval in the UV range: an application to OMI satellite validation. Atmos. Chem. Phys. Discuss. 9, 19009–19033. Jorquera, H., 2002. Air quality at Santiago, Chile: a box modeling approach. I. Carbon monoxide, nitrogen oxides and sulfur dioxide. Atmos. Environ. 36, 315–330. Kazadzis, S., Bais, A., Balis, D., Kouremeti, N., Zempila, M., Arola, A., Giannakaki, E., Amiridis, V., Kazantzidis, A., 2009a. Spatial and temporal UV irradiance and aerosol variability within the area of an OMI satellite pixel. Atmos. Chem. Phys. 9, 4593–4601. Kazadzis, S., Bais, A., Arola, A., Krotkov, N., Kouremeti, N., Meleti, C., 2009b. Ozone Monitoring Instrument spectral UV irradiance products: comparison with ground based measurements at an urban environment. Atmos. Chem. Phys. 9, 585–594. Krotkov, N.A., Barthia, P.K., Herman, J.R., Fioletov, V., Kerr, J., 1998. Satellite estimation of spectral surface UV irradiance in the presence of tropospheric aerosols. J. Geophys. Res. 103, 8779–8793. http://dx.doi.org/10.1029/ 98JD00233.

291

Levy, R., Remer, L., Mattoo, S., Vermote, E., Kaufman, Y., 2007. Secondgeneration operational algorithm: retrieval of aerosol properties over land from inversion of Moderate Resolution Imaging Spectroradiometer spectral reflectance. J. Geophys. Res. 112, D13. http://dx.doi.org/10. 1029/2006JD007811. Levy, R., Remer, L., Kleidman, R., Mattoo, S., Ichoku, C., Kahn, R., Eck, T., 2010. Global evaluation of the Collection 5 MODIS dark-target aerosol products over land. Atmos. Chem. Phys. 10, 10399–10420. Mayer, B., Kylling, A., 2005. Technical note: the libRadtran software package for radiative transfer calculations — description and examples of use. Atmos. Chem. Phys. 2005 (5), 1855–1877. http://dx.doi.org/10.5194/acp-5-18552005. McKenzie, R.L., Bodeker, G.E., Scott, G., Slusser, J., 2006. Geographical differences in erythemally-weighted UV measured at mid-latitude USDA sites. Photochem. Photobiol. Sci. 5, 343–352. McKenzie, R., Weinreis, C., Johnston, C., Liley, B., Shiona, H., Kotkamp, M., Smale, D., Takegawa, N., Kondo, Y., 2008. Effects of urban pollution on UV spectral irradiances. Atmos. Chem. Phys. Discuss. 8, 7149–7188. Mei, L., Xue, Y., Xu, H., Guang, J., Li, Y., Wang, Y., Ai, J., Jiang, S., He, X., 2012. Validation and analysis of aerosol optical thickness retrieval over land. Int. J. Remote Sens. 33, 781–803. Münkel, C., Eresmaa, N., Räsänen, J., Karppinen, A., 2007. Retrieval of mixing height and dust concentration with lidar ceilometer. Boundary Layer Meteorol. 124, 117–128. Muñoz, R.C., Alcafuz, R., 2012. Variability of urban aerosols over Santiago, Chile: comparison of surface PM10 concentrations and remote sensing with ceilometer and lidar. Aerosol Air Qual. Res. 12, 8–19. Muñoz, R.C., Undurraga, A., 2010. Daytime mixed layer over the Santiago Basin: description of two years of observations with a lidar ceilometer. J. Appl. Meteorol. Climatol. 49, 1728–1741. Oyanadel, A., Painemal, D., Leon, J.-F., Chiapello, I., Gallardo, L., 2006. Aerosol loading over Santiago De Chile (33°.3′S 70°.5′W, 500 M.A.S.L.): a comparison between satellite and in situ measurements. Proceedings of 8 ICSHMO, INPE 149–155. Perez, P., Salini, G.A., 2008. PM2.5 forecasting in a large city: comparison of three methods. Atmos. Environ. 42, 8219–8224. Ruiz-Arias, J.A., Dudhia, J., Gueymard, C.A., Pozo-Vázquez, D., 2013. Assessment of the Level-3 MODIS daily aerosol optical depth in the context of surface solar radiation and numerical weather modeling. Atmos. Chem. Phys. 13, 675–692. http://dx.doi.org/10.5194/acp-13-675-2013. Rutllant, J., Garreaud, R., 1995. Meteorological air pollution potential for Santiago, Chile: towards an objective episode forecasting. Environ. Monit. Assess. 34, 223–244. Schaap, M., Apituley, A., Timmermans, R.M.A., Koelemeijer, R.B.A., de Leeuw, G., 2009. Exploring the relation between aerosol optical depth and PM2. 5 at Cabauw, The Netherlands. Atmos. Chem. Phys. 9, 909–925. Seckmeyer, G., Bais, A., Bernhard, G., Blumthaler, M., Booth, C.R., 2001. Disterhoft P, Eriksen P, McKenzie R L, Miyauchi M and Roy C Part 1: Spectral instruments. Instruments to Measure Solar Ultraviolet Radiation WMOGAW No. 125. World Meteorological Organization, Geneva, Switzerland. Stamnes, K., Slusser, J., Bowen, M., 1991. Derivation of total ozone abundance and cloud effects from spectral irradiance measurements. Appl. Opt. 30, 4418–4426. Tanskanen, A., Krotkov, N.A., Herman, J.R., Arola, A., 2006. Surface ultraviolet irradiance from OMI. IEEE Trans. Geosci. Remote Sens. 44 (5), 1267–1271. Tanskanen, A., Lindfors, A., Määttä, A., Krotkov, N., Herman, J., Kaurola, J., Koskela, T., Lakkala, K., Fioletov, V., Bernhard, G., McKenzie, R., Kondo, Y., O'Neill, M., Slaper, H., Den Outer, P., Bais, A., Tamminen, J., 2007. Validation of daily erythemal doses from Ozone Monitoring Instrument with ground-based UV measurement data. J. Geophys. Res. 112. http:// dx.doi.org/10.1029/2007JD008830 (D24S44). Thomas, C., Briottet, X., Santer, R., Lacherade, S., 2008. Aerosols in urban areas: optical properties and impact on the signal incident to an airborne high spatial resolution camera. Proc. SPIE 7107, 71070O. Torres, O., Tanskanen, A., Veihelman, B., Ahn, C., Braak, R., Bhartia, P., Veefkind, P., Levelt, P., 2007. Aerosols and surface UV products from OMI observations: an overview. J. Geophys. Res. 112. http://dx.doi.org/10.1029/2007JD008809 (D24S47). Tsai, Y., Jeng, Y., Chu, A., Chen, J., Chang, S., 2011. Analysis of the relationship between MODIS aerosol optical depth and particulate matter from 2006 to 2008. Atmos. Environ. 45, 4777–4788. Weihs, P., Blumthaler, M., Rieder, H., Kreuter, A., Simic, S., Laube, W., Schmalwieser, A.W., Wagner, J.E., Tanskanen, A., 2008. Measurements of UV irradiance within the area of one satellite pixel. Atmos. Chem. Phys. 8, 5615–5626. Wuttke, S., Seckmeyer, G., Bernhard, G., Ehramjian, J., McKenzie, R., Johnston, P., O'Neill, M., 2006. New spectroradiometers complying with the NDSC standards. J. Atmos. Ocean. Technol. 23 (2), 241–251.