Remote sensing measurements of the volcanic ash plume over Poland in April 2010

Atmospheric Environment 48 (2012) 66e75

Contents lists available at ScienceDirect

Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv

Remote sensing measurements of the volcanic ash plume over Poland in April 2010 K.M. Markowicz a, *, T. Zielinski b, A. Pietruczuk c, M. Posyniak a, O. Zawadzka a, P. Makuch b, I.S. Stachlewska a, A.K. Jagodnicka d, T. Petelski b, W. Kumala a, P. Sobolewski c, T. Stacewicz d a

Institute of Geophysics, Faculty of Physics, University of Warsaw, ul. Pasteura 7, 02-093 Warsaw, Poland Institute of Oceanology PAS, ul. Powst. W-wy 55, 81-712 Sopot, Poland Institute of Geophysics PAS, ul. Ksiecia Janusza 64, 01-452 Warsaw, Poland d Institute of Experimental Physics, Faculty of Physics, University of Warsaw, ul. Hoza 69, 00-681 Warsaw, Poland b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 April 2011 Received in revised form 16 June 2011 Accepted 5 July 2011

This work provides information on selected optical parameters related to volcanic ash produced during the eruption of the Eyjafjöll volcano in Iceland in 2010. The observations were made between 16 and 18 April 2010 at four stations representative for northern (Sopot), central (Warsaw, Belsk) and south-eastern (Strzyzow) regions of Poland. The largest ash plume (in terms of aerosol optical thickness) over Poland was observed at night of 16/17 April 2010 in the layer between 4 and 5.5 km a.s.l. The highest values of the aerosol extinction coefficient reached 0.06e0.08 km
1 at 532 nm (based on lidar observations in Warsaw) and 0.02e0.04 km
1 at 1064 nm (based on ceilometer observations in Warsaw). The corresponding optical thickness due to volcanic ash reached values of about 0.05 at 532 nm and about 0.03 at 1064 nm. These values are similar to those reported for the Belsk station based on lidar observations. The ash mass concentration estimated based on the maximum aerosol extinction coefficient reached 0.22  0.11 mg m
3. This value is significantly lower than the limit (2 mg m
3) for the aircraft operation. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: Aerosol optical thickness Remote sensing Volcano ash Retrieval Ceilometer Lidar Sun photometer

1. Introduction Volcanoes are among most important natural sources of atmospheric pollution. They emit dust and gases, which undergo chemical reactions in the atmosphere producing aerosols (McCormick et al., 1995). Similar to desert dust, particles produced as a result of fires or industrial combustion processes, volcanic aerosols may influence meteorological conditions e.g. surface solar radiation, visibility, and other. For example the Tambora volcano eruption in April 1815 caused a year with no summer in Indonesia in 1816 (Trigo et al., 2009). The climatic impact of volcanic eruptions is usually less spectacular, however, very important and can also be observed (Hansen et al., 1992; Parker et al., 1996; Kirchner et al., 1999). Particle size distribution of volcanic aerosols tends to be multimodal, suggesting multiple processes of formation. The finest particles result from the condensation of volatiles, and gas phase reactions (Mather et al., 2003). The accumulation mode includes also the smallest tephra particles, while the coarse mode includes

* Corresponding author. Tel.: þ4822 5546836; fax: þ4822 5546882. E-mail address: [email protected] (K.M. Markowicz). 1352-2310/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2011.07.015

fragmented magma, and erosion particles (Mather et al., 2003). Impact of volcanoes on global climate results from a relatively long life of fine particles in the atmosphere. The residence time of fine particles (mainly sulfuric compounds) in the stratosphere varies in a wide range from months to years. In case of volcanic ash, the climate effect is usually smaller, because particles have a maximum residence time in the troposphere of few weeks. Only the finest tephra particles remain in the stratosphere for up to few months, but they have only minor climatic effect. Volcanic aerosols mainly scatter solar radiation, however, small absorption rate which occurs in upper atmospheric layers has an important impact on local radiation balance (Myhre et al., 2001; Harshvardhan, 1979) and as a consequence, on air temperature (Parker et al., 1996; Kirchner et al., 1999). An indirect influence of volcanic aerosols on climate is related to their impact on microphysical cloud properties (Durant et al., 2008). The motivation for this study was a several day long closing of the European air space (IATA, 2010) after the eruption of the Eyjafjöll volcano in April 2010. Such situation resulted from the lack of consistent volcanic ash monitoring over Europe. This work provides information on spatial and temporal evolution of the aerosol optical properties such as aerosol optical thickness (AOT) and aerosol extinction and backscatter coefficients related to

K.M. Markowicz et al. / Atmospheric Environment 48 (2012) 66e75

volcanic ash produced during the eruption of the Eyjafjöll volcano which was observed over Poland between 16 and 18 April 2010. It is worth mentioning that other volcanic ash episodes, not reported in this work, were observed over central Poland as well (Pietruczuk et al., 2010; Mona et al., 2010; Campanelli et al., 2012; Thomas and Prata, 2010). 2. Methods 2.1. Instrumentation The volcanic ash optical properties were measured with the use of the remote techniques at 4 stations in Poland (Fig. 1). The EUSAAR Sopot station is based in the Institute of Oceanology Polish Academy of Sciences (www.iopan.gda.pl) on the coast of the Baltic Sea (54 260 N, 18 330 E, 2 m a.s.l.). The laboratory is equipped with Microtops II sunphotometers (5 wavelengths), lidar, nephelometer, and particle counters. For the purpose of this paper only sun photometer data were used (Table 1). The University of Warsaw Radiative Transfer Laboratory (http:// www.igf.fuw.edu.pl) is based on the platform on the roof of the university building (52 210 N, 20 980 E, 110 m a.s.l.). The instrumentation involves a CHM-15K ceilometer made by JenOptik and a Microtops II sunphotometer. Additionally, a 510M lidar of the Faculty of Physics of the University of Warsaw was employed for measurements during the measurement period. This lidar is based on the NdeYAG laser generating beams at its three harmonics: 1064, 532 and 355 nm, where energies of the light pulses are 170, 90 and 60 mJ, respectively. During the Eyjafjoll volcanic event setup of the 510M lidar contained only 532 nm wavelength. The measurements were taken with pulse repetition frequency of 10 Hz. The Geophysical Observatory in Belsk is equipped with an aerosol backscatter LIDAR and a Sun-scanning photometer. The Belsk LIDAR, a part of EARLINET network, has a Nd:YAG laser with three harmonics. Aerosol backscatter coefficients at wavelengths 1064, 532 and 355 nm are calculated by means of a standard Fernald procedure and are available through the EARLINET database (http://earlinet.org). Collocated Sun-photometers CIMEL CE 318, is federated in European SkyRad users’ network (ESR) and AErosol RObotic NETwork (AERONET) networks respectively. The EUSAAR SolarAOT private station (established by K. Markowicz in 2003 http://www.igf.fuw.edu.pl/meteo/stacja/) in Strzyzow,

Fig. 1. Location of Polish stations.

67

in the south-east part of Poland (49 860 N, 21 870 E, 443 m a.s.l.) uses a Multi-Filter Rotating Shadowband Radiometer (Model MFR-7) (Harrison et al., 1994). The volcanic ash analyses were supported with satellite data from (Moderate Resolution Imaging Spectroradiometer) (MODIS) mounted on Terra and Aqua satellite and Spinning Enhanced Visible and Infra-red Imager (SEVIRI) onboard of theMeteosat Second Generation (MSG). The authors analyzed data provided by the AErosol RObotic NETwork (AERONET) at different stations placed on the route of the volcanic ash plume. In addition, we used the Lagrangian particle dispersion model FLEXPART (Stohl et al., 1998) and the Hybrid Single Particle Lagrangian Integrated Trajectory Model (HYSPLIT) (Draxler and Rolph, 2010). 2.2. Retrieval techniques To obtain vertical profiles of the aerosol extinction from ceilometer signals we used three different retrieval methods (Frey et al., 2010; Heese et al., 2010; Flentje et al., 2010; Markowicz et al., 2008). These require additional information about aerosol optical properties such as a lidar ratio (ratio of particle extinction to backscatter coefficient) and/or aerosol optical thickness (AOT). The first method used in this study, called the Porter approach (Porter et al., 2000, revised by Markowicz et al., 2008) is based on the forward-stepping algorithm which requires information about the single scattering albedo and the backward phase function or the lidar ratio, as well as the lidar calibration coefficient. The latter quantity was obtained using the lidar auto-calibration technique (O’Connor et al., 2004), which we modified for 1064 nm. This method had to be applied to signals detecting a persistent low level and not too-thick Cumulus cloud system, which causes complete attenuation of the ceilometer signal but without causing its saturation. Another Porter method (Porter et al., 2002) does not require knowledge of the lidar calibration coefficient (because it is constrained using sun photometer measurements) but we did not use this method. The two other methods applied to the ceilometer signals were the standard backward KletteFernaldeSasano approach (Klett, 1985; Fernald, 1984; Sasano et al., 1985) and the seldom used forward KletteFernaldeSasano approach. For the three evaluation approaches the lidar ratio was assumed constant with altitude. However, for the KletteFernaldeSassano approach the lidar ratio was calculated as an adjustment of an integrated ceilometer extinction coefficient profile to the total AOT measured by the sun photometer. By using additional information, such as the value of the integrated aerosol optical thickness from sun photometer, the KletteFernaldeSasano technique, can reduce the error of the derived extinction profiles to 0.005 km
1 (Welton and Campbell, 2002). On the other hand an assumption of the lidar ratio constant with altitude may lead to significant uncertainties (Sasano et al., 1985). In case of the Porter approach we assumed the lidar ratio of 50 sr constant with altitude and time, a value typical for volcanic ash particles (Wang et al., 2008; Papparlardo et al., 2010; Ansmann et al., 2010). In the case of the two forward approaches the initial aerosol extinction coefficient (at 0.25 km) was assumed 0.01 km
1 at 1064 nm. However, during the consecutive iterations this value varied, so that the final aerosol extinction coefficient at 0.25 km may differ from the starting value. Assuming an error of 2% of the molecular extinction coefficient calculated from the radiosounding data, which is accounted to a daily variation of temperature and pressure, the error of the retrieved aerosol extinction coefficient is about 10% (Stachlewska and Ritter, 2010). As the total optical thickness measurements with sun photometers are not possible at night the AOT used for the aerosol extinction coefficient retrieval was obtained by the interpolation of

68

K.M. Markowicz et al. / Atmospheric Environment 48 (2012) 66e75

Table 1 List of instrumentation used in this study and measured quantities a e Ǻngstrom exponent, s(z), b(z) e profile of the aerosol extinction and backscatter coefficients. Stations

Photometers

Sopot Warsaw

Four Microtops (340, 380, 400, 500, 675, 870, 1020 nm) Microtops 380, 500, 675, 870, 1020 nm)

Belsk Strzyzow

CIMEL (340, 380, 440, 500, 670, 870, 940, 1020 nm) MFR-7 (415, 500, 605, 675, 870, 940 nm)

the data collected in the early morning and afternoon of the following day. The different optical path for the remote zenith-pointing ceilometer and the solar photometer pointing towards the Sun results in the error which has been estimated up to 10% for the polar region, where these differences are the largest (Stachlewska and Ritter, 2010). The errors resulting from the assumption of the overlap correction function remained unchanged with time between 0.25 and 0.6 km and are less than 3% (Stachlewska and Markowicz, 2010). The aerosol extinction profiles from the 510M lidar are estimated from the standard backward KletteFernaldeSasano technique. In case of Belsk measurements extinction was obtained from aerosol backscatter coefficient profiles stored in EARLINET database multiplied by constant lidar ratio of 50 sr, which is in accordance with that observed in Belsk using a sun photometer (Pietruczuk and Podgórski, 2009). The volcanic ash analyses were supported with satellite data from the SEVIRI/MSG. We developed an algorithm to detect volcanic aerosol based on the EUMETSAT (2007) report. It allows to indentify ash layers above clouds. Ash particles above ocean or land surface are not detected by this technique. We constructed the RGB picture based on the following methodology. Channel R is defined by a ratio of the reflectance measured at 3.9 mm and 0.6 mm, channel G by the brightness temperature difference between 8.7 mm and 10.8 mm, and channel B as the brightness temperature difference between 12.0 mm and 10.8 mm. This technique exploits the fact, that the absorption of clouds (ice and water) is higher at 12.0 mm than at 10.8 mm, but it is just the opposite for ash plumes (Ellrod et al., 2003). In addition, the single scattering albedo of water and ice clouds decreases and for ash particles increases with wavelength in the visible and near infrared (Pavolonis et al., 2006). Therefore, the ratio of the solar contribution in channel 3.9 mm with the visible channel 0.6 mm, ash can be separated from ice or water clouds. The last relationship (for channel B) uses the fact, that SO2 has a maximum absorption close to 9 mm (Pavolonis et al., 2006). In such procedure the ash clouds are yellow or orange in a constructed picture. 3. Observations During the eruption of the Eyjafjöll volcano (14 April 2010) there was a high pressure system over western and central Europe with a center north-west of the British Isles and a vast low pressure system over the north of Scandinavia. The geopotential field at 500 hPa near Iceland was latitudally oriented causing transportation of air masses towards Scandinavia. In the next two days the high pressure system moved east turning the geopotential field in the middle troposphere. The wind direction at 500 hPa in the North Sea region was north-westerly. Similar situation was at a ground level. During 16 of April the cold front associated with the low pressure system (located over the Norwegian Sea) was passing the Southern Baltic Sea and Poland in a south-easterly direction. Backtrajectories obtained from the HYSPLIT (Fig. 2 a, b) indicate

Lidars 510M Lidar (532, 1064 nm) ceilometer CHM-15K (1064 nm) Lidar (355, 532, 1064 nm)

Computed quantities AOT, a AOT, a, s(z), b(z) AOT, a, s(z), b(z) AOT, a

continental air masses from southern Europe preceding this cold front. Polar marine air masses followed the front. Simulations of the dispersion of volcanic ash by the Lagrangian particle dispersion model FLEXPART (Stohl et al., 2011) show (Fig. 2) that this air mass contains volcanic particles and is moving in the south-east direction. The ash plume passed Sopot in the morning, Warsaw and Belsk in the afternoon of 16 April, and Strzyzow at midnight of 17 April. The maximum of ash concentration in the whole column computed by this model exceed 2000 mg m
2. The backtrajectories ending at 12:00 UTC on 17 April (Fig. 2c) show that the air masses which reached Polish stations, around 30 h earlier were still over Iceland. The trajectories at different altitudes do not differ significantly in terms of the type of air mass carried. Next day backtrajectories did not exactly cross over Iceland (Fig. 2d) but transport of volcanic aerosols cannot be excluded due to model uncertainty and particle dispersion around the Iceland. However, the FLEXPART does not predict the ash particles over Poland on 17 and 18 of April. Similarly to the dispersion model the satellite data from MSG show ash cloud passing over Poland on 16 April (Fig. 4). At 06:12 UTC the ash particles (yellow color) are observed at the Polish coast just before the Gulf of Gdansk and at 16:12 UTC they are visible over Warsaw. Position of the ash cloud from the MSG data is consistent with the FLEXPART results. Fig. 5 shows a logarithm of range and overlap corrected CHM15K ceilometer signal measured in Warsaw between 16 and 18 April 2010. On 16 April ceilometer data show large backscatter signals from cloud layers between 2 and 3 km a.s.l. Only close to midnight, sky was not obscured by clouds which allowed upper troposphere measurements. Aerosols were present only in the Planetary Boundary Layer (PBL), however LIDAR measurements performed in Belsk show presence of weak aerosol layers (between 3 and 5 km and between 8 and 9 km a.s.l.) at the end of 16 April (Fig. 6). That layers are still visible on 17 April in the morning which is consistent with ceilometer results showing a plume of aerosols up to 5.5 km a.s.l. from 00:00 UTC on 17 April. Identification of the volcanic aerosol without chemical analysis or information about lidar depolarization is difficult. However, based on seven years observations performed by the CT-25K (Vaisala) and CHM-15K ceilometers (Markowicz et al., 2008) we have concluded that only in case of desert dust event this instrument is able to detect aerosol layers in the middle troposphere (about 3e7 km a.s.l.). These events are observed mainly during south and south-westerly circulation which during that volcano episode did not occur. Thus, the observed plum can be identified as ash particles. In addition, data obtained from HYSPLIT (Fig. 2), FLEXPART (Fig. 3) and MSG (Fig. 4) support such interpretation. In addition, the FLEXPART simulation (not shown) predicted the ash plume between 2 and 6 km with a maximum close to 5 km, which is consistent with lidar and ceilometer observations. The next day (17 April) the middle troposphere plume disappeared because of shading cumulus cloud effect and probably because of poor signal to noise ratio of ceilometer during daytime. However, the aerosol plume was detected by Belsk LIDAR in the gaps between clouds (Fig. 6). On the night of 17/18 April slight aerosol layer above the PBL was detected.

K.M. Markowicz et al. / Atmospheric Environment 48 (2012) 66e75

69

Fig. 2. HYSPLIT backtrajectories obtained for Sopot, Warsaw and Strzyzow ending at 00 UTC on 16 April (a), 06 UTC on 16 April (b), 12 UTC on 17 April (c), and 12 UTC on 18 April 2010 (d).

Aerosol extinction coefficients derived from the ceilometer and the 510M lidar measurements performed in Warsaw as well as measurements taken in Belsk show a plume of aerosols descending with time in the middle troposphere (Figs. 7 and 8). The soliddotted line, dotted line and solid line depict the aerosol extinction coefficient profiles obtained from the ceilometer signals at 1064 nm using the Porter approach, and the backward and forward KletteFernaldeSasano approaches, respectively. The solid line with squares depicts the aerosol extinction coefficient profiles obtained from the lidar signals at 532 nm using the backward Klette FernaldeSasano. The aerosol extinction coefficients derived from ceilometer measurements show plumes of aerosols in the middle troposphere which were descending with time (Fig. 7). Solid-dotted line, dotted line and solid line show the aerosol extinction coefficients obtained

from the Porter approach (Porter et al., 2000), backward Klette FernaldeSasan (Klett, 1985; Fernald, 1984; Sasano et al., 1985), and forward KletteFernaldeSasano, respectively. These methods are commonly used to retrieve the aerosol extinction coefficient from ceilometer observations (Markowicz et al., 2008; Heese et al., 2010; Flentje et al., 2010), while different methods provide some information about the aerosol extinction uncertainties. We found that the variability of the volcanic extinction coefficient with different methods is about 30e40%. On 17 April at midnight the plume of ash was detected in a layer from 4.5 to 5.5 km a.s.l. and 90 min later the plume was observed in a layer between 4 and 5 km a.s.l. The highest observed values of the aerosol extinction coefficient are between 0.02 and 0.04 km
1 at 1064 nm, depending on the approach applied for the retrieval from the ceilometer data. For the lidar the highest observed value vary

70

K.M. Markowicz et al. / Atmospheric Environment 48 (2012) 66e75

Fig. 3. Total column volcanic ash mass concentration [mg m
2] between 16 April at 06:00 UTC and 18 April at 12:00 UTC obtained from the FLEXPART model.

between 0.06 and 0.08 km
1 at 532 nm. Relatively smaller values were observed in Belsk (Fig. 9). The aerosol extinction coefficient does not exceed 0.02 km
1 at both (532 and 1064 nm) wavelengths. In addition the spectral difference of the extinction coefficients in

the plume layers was small, which indicates presence of coarse mode particles. The estimated ash optical thickness using the integration of retrieved extinction profiles was not larger than 0.03 at 1064 nm

Fig. 4. The RGB composition of SEVIRI/MSG data at 06:12 UTC (a), 12:12 UTC (b), 16:12 UTC (c) on 16 April and at 12:12 UTC on 17 April. Yellow and orange colors correspond to volcanic ash, blue and red color mark ice and water cloud, respectively, brown and black color denote clear sky surface.

K.M. Markowicz et al. / Atmospheric Environment 48 (2012) 66e75

Fig. 5. Logarithm of range and overlap corrected CHM-15K ceilometer signal measured in Warsaw between 16 and 19 April 2010.

and 0.05 at 532 nm. It corresponds with the values of optical thicknesses obtained in Belsk by Pietruczuk et al. (2010). Twenty hours later the aerosol extinction in the middle troposphere was smaller (in order of 0.01 km
1). In the PBL the plume was thicker and was observed in a layer from 600 m to 2 km a.s.l. and it was even lower on 18 April at 1 a.m. However, the PBL aerosols are probably mainly anthropogenic. Interesting information comes from the analyses of aerosol optical measurements made with Microtops II sun photometer, MFR-7 shadowband and CIMEL sun photometer at four locations in Poland (Fig. 10). Between 16 and 19 April the AOT values at 500 nm registered at 4 stations in Poland were above 0.1. The highest values

71

were recorded on 16 April in Warsaw (0.23 up to 0.45) and in Belsk (0.23 up to 0.43), although increase of coarse aerosol fraction, which could be dust, was observed on 17 April in Belsk (Pietruczuk et al., 2010). The AOT measured in Sopot was much lower, ranging from 0.18 to 0.22. However, there were very few good measurements made on that day due to high cloudiness in the region. Large values of the AOT measured in Warsaw and Belsk on 16 April are related to continental polluted air masses (Fig. 2a, b) and not to the volcanic event. Only increased concentration of coarse mode registered on 17 April could be connected with the volcanic event. In Sopot observations of the AOT were performed after the cold front passage (around 6 UTC) and therefore the AOT values were significantly lower. Generally, advection of air masses from the North Atlantic leads to the reduction of the AOT. Typical value of the AOT for this circulation is about 0.1e0.15 (at 500 nm) (Maciszewska et al., 2010). During the next two days the AOT values were decreasing and ranged from 0.12 to 0.2. Over a course of 3 days the AOT values were decreasing latitudinally. Interestingly the Ǻngstrom exponent values from the AOT at 500 and 1020 nm, show contribution of larger particles observed in Sopot on 16 April in comparison with the situation in Warsaw and Belsk, where during following two days the values were similar. However, on 18 April the Ǻngstrom exponent values derived for the furthest south station in Strzyzow are higher than in Warsaw and in Belsk which indicates the presence of fine particles. The AOT values at 550 nm obtained from MODIS for a period between 16 and 18 April are presented in Fig. 11. It is obvious that on 16 April the significant cloud coverage of the entire region made the satellite-based data collection impossible. This picture explains “why the data obtained using the ground-based techniques are scarce and could be collected only during episodes of clear sky, windows” between clouds. The MODIS values are comparable with those obtained by the authors for Warsaw. The situation slightly improved on 17 April, however, there are no MODIS data for the stations in Warsaw and Strzyzow. On 18 April the ground-based

Fig. 6. Logarithm of attenuated aerosol backscattering coefficient (range corrected signal) measured in Belsk on 16 (a) and 17 (b, c) April 2010.

72

K.M. Markowicz et al. / Atmospheric Environment 48 (2012) 66e75

Fig. 7. Aerosol extinction coefficient obtained from the CHM-15K ceilometer signals at 1064 nm and the 510M lidar at 532 nm, both calibrated using sun photometer observations in Warsaw. The ceilometer profiles were calculated using the Porter approach (solid-dotted line), the backward KletteFernaldeSasano approach (dotted line), the forward KletteFernaldeSasano (solid line) approach. The lidar profiles were obtained using the backward KletteFernaldeSasano (squares with line) method. Three panels correspond to data averaged between 23:30 and 00:00 UTC on 16 April 2010 (a), between 00:00 and 00:30 UTC on 17 April 2010 (b), and between 01:30 and 02:00 UTC (c) on 17 April 2010.

values of the AOT and the satellite ones are in good agreement. MODIS data show reduction of the AOT during the change of the air mass advection between 16 and 17 April. After the inflow from Iceland the AOT was significantly lower which indicates small effect

of volcanic ash on the total AOT. These results agree with the ceilometer observations over Warsaw, where the estimated ash AOT at 1064 nm was only 0.03. In addition, spectral dependence of the single scattering albedo retrieved from the CIMEL almucantar scans

Fig. 8. Aerosol extinction coefficient obtained by CHM-15K ceilometer (in Warsaw) at 1064 nm and sun photometer observation using Porter approach (solid-dotted line), backward KletteFernaldeSasano (dotted line), forward KletteFernaldeSasano (solid line) method. Panels correspond to data averaged between 20:00 and 20:30 UTC on 17 April 2010 and between 00:45 and 01:15 UTC on 18 April 2010 (b).

K.M. Markowicz et al. / Atmospheric Environment 48 (2012) 66e75

73

Fig. 9. Aerosol extinction coefficient [km
1] derived from aerosol backscatter coefficient measured in Belsk on 16 and 17 April 2010.

in Belsk (Fig. 12) shows typical (climatologic) behavior. The single scattering albedo decreases with wavelength for all observations performed between 16 and 18 of April 2010. However, the slope of the spectral variability is slightly smaller in comparison to the 9 year (2002e2010) averaged data (solid line). The different slope can be explained by the presence of the ash particles in the atmosphere. The single scattering albedo of volcanic aerosol increases with wavelength (Lindqvist et al., 2011) as well as the mineral dust (Li et al., 2007). Because we did not observe such spectral variability but only reduction of the opposite slope, this indicates relatively small contribution of the volcanic AOT to the total AOT. Single

Fig. 10. Temporal variability of the aerosol optical thickness at 500 nm (a) and the Ǻngstrom exponent a ¼
ln(s500/s1020)/n(l500/l1020) (b) measured with Microtops II sun photometer in Warsaw (open circles), in Sopot (open squares), with CIMEL in Besk (open diamonds), and with MFR-7 in Strzyzow (solid dots) between 16 and 18 April 2010.

Fig. 11. Aerosol optical thickness at 550 nm obtained from MODIS measurement onboard the Terra and Aqua satellite. Pictures (a, b, c) correspond to overpasses on 16, 17 and 18 April 2010, respectively, which took place between 9 and 11:30 UTC. The data from level 2 of the MOD04_L2 and MYD04_L2 product of collection 051 have been used.

74

K.M. Markowicz et al. / Atmospheric Environment 48 (2012) 66e75

Fig. 12. Spectral variability of the single scattering albedo retrieved from CIMEL sun photometer in Belsk (AERONET data level 2.0) between 16 and 18 April 2010. The solid line corresponds to April data averaged between 2002 and 2010.

scattering albedo, except the one result on 16 April, is relatively low (between 0.86 and 0.92 at 441 nm). 4. Discussion and conclusions The results presented in this paper show clearly that due to high cloud coverage the period between 16 and 18 April 2010 was difficult for remote sensing of the atmosphere in the region of central Europe. This is especially evident on 16 April. There are only very few data available for this day from both ground-based and satellite measurements. The AOT values were decreasing latitudinally and the Ǻngstrom exponent values show higher presence of fine mode particles on 18 April at the station in Strzyzow than in Warsaw. The AOT values ranged from 0.12 to 0.4 between 16 and 18 April and were the highest on 16 April. However, that day was also most difficult in terms of measurements. Our observations do not agree very well with those reported by Ansmann et al. (2010). Ansmann et al. stated that the AOT measured in Leipzig reached 1.0 and the ash-related optical thickness was about 0.7. However, in the capture to Fig. 4 the authors explain that these results (16 April 2010) are based on the AERONET data level 1.0. Data from this level are not cloud-screened. A thorough check of the data provided by the AERONET at different stations in the central Europe region for 16 April show very little or lack of data of levels 1.5 (cloudscreened) and level 2.0 (quality-assured). And thus, the station in Hamburg (53 N 9 E) provides very few data points level 1.5 for the 16 of April. The average AOT values at 500 nm are 0.31. No data level 2 are available. Station in Helgoland (54 N 7 E) provides some data of level 1.5 for 16 April, however, also not too many. The average AOT at 500 nm is 0.21. No level 2 data. The station in Leipzig (51 N 12 E) provides no data level 2 and only 2 (!) measurement points level 1.5 after 164:00 UTC with an averaged AOT at 500 nm of 0.34. The indication is obvious that clouds interrupted the measurements, and the data, if available, are provided only after 16 UTC. Ansmann et al. reported that “The Leipzig AERONET photometer registered peak AOT around 1.0 and asherelated values up to 0.7 at 500 nm from 13:20e13:40 UTC on 16 April 2010”. They did not explain, however, how the value 0.7 for ash was estimated from the AERONET data. Similar as the total optical thickness this value is affected by clouds. On 17 April AERONET data from level 2.0 show

AOT at 500 nm variability between 0.15 and 0.2. Similar values (0.11e0.16) were measured in Warsaw (about 650 km from Leipzig) in Poland and at the AERONET station in Belsk (0.12e0.2 data level 2.0), located 40 km south-east of Warsaw. Estimation of volcanic ash mass concentrations from extinction or backscatter profiles is difficult because, it requires information about aerosol size distribution, mass density as well as the refractive index and particle shape. To simplify this problem we used the particle properties of desert dust according to the Optical Properties of Aerosol and Clouds database (Hess et al., 1998). The mineral transported category describes desert dust that is transported over long distances with a reduced amount of large particles. It allows to compute extinction to mass conversion factor which is about 2700 mg m
3km
1 for 532 nm. Thus, a maximum of the aerosol extinction coefficient measured during the volcanic event of 0.08 km
1 corresponds to mass concentration of 0.22 mg m
3. Uncertainty of this estimation of the mass concentration is about 35% (Ansmann et al., 2010). In addition, significant source of uncertainty of the mass concentration is related to the uncertainty of the aerosol extinction coefficient. We found that variability of the maximum of the volcanic extinction coefficient with different inversion approaches is about 30e40%. It leads to the total mass concentration uncertainty of about 50%. Making an assumption that the mass concentrations reported by Ansmann et al., in order of 1 mg m
3, the results reported by Flentje et al. (2010) from lidar and sun photometer observations at 0.5e0.75 mg m
3, and by Gasteiger et al. (2010) maximum mass concentration of 1.1 mg m
3 (0.65e1.7 mg m
3), were properly measured they are still below the legal flight threshold of 2 mg m
3. The European Commission (2010) defined following ash concentration thresholds: less than 0.2 mg m
3 for the “Normal” flying zone, between 0.2 and 2 mg m
3 for “No-intentional flight” zone, and more than 4 mg m
3 for “No-Fly” zone. Stohl et al. (2011) found, based on the Lagrangian dispersion model from satellite data, that during three episodes in April and May 2010, volcanic ash concentrations at some altitudes exceeded the limits for the “Normal” flying zone up to 14%, 2% and 7%, respectively. For a limit of 2 mg m
3 only two episodes with fractions of 1.5% and 0.9% occurred, while the current “No-Fly” zone was rarely exceeded. Acknowledgments This research has been partly made within the Polish National Grants POLAR-AOD, No. NN 306315536 coordinated by the IO PAS, grants No. 1283/B/P01/2010/38 and No. 1276/B/P01/2010/38 of the Ministry of Science and Higher Education of Poland, both coordinated by IGF UW. We acknowledge Brent Holben for the use of the data from the AERONET station in Belsk and EUMETSAT for the data availability, license number 50001643. References Ansmann, A., Tesche, M., Gross, S., Freudenthaler, V., Seifert, P., Hiebsch, A., Schmidt, J., Wandinger, U., Mattis, I., Wiegner, M., 2010. The 16 April 2010 major volcanic ash plume over central Europe: EARLINET lidar and AERONET photometer observations at Leipzig and Munich, Germany. Geophys. Res. Lett. 37 (L13810). doi:10.1029/2010GL043809. Campanelli, M., Estelles, V., Smyth, T., Tomasi, C., Martìnez-Lozano, M.P., Claxton, B., Muller, P., Pappalardo, G., Pietruczuk, A., Shanklin, J., Colwell, S., Wrench, C., Lupi, A., Mazzola, M., Lanconelli, C., Vitale, V., Congeduti, F., Dionisi, D., Cardillo, F., Cacciani, M., Casasanta, G., 2012. Monitoring of Eyjafjallajökull volcanic aerosol by the new European Skynet Radiometers (ESR) network. Atmos. Environ 48, 33e45. Draxler, R.R., Rolph, G.D., 2010. HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory). NOAA Air Resources Laboratory, Silver Spring, MD. Model access via NOAA ARL READY Website. http://ready.arl.noaa.gov/HYSPLIT.php. Durant, A.J., Shaw, R.A., Rose, W.I., Mi, Y., Ernst, G.G.J., 2008. Ice nucleation and overseeding of ice in volcanic clouds. J. Geophys. Res. 113 (D09206). doi:10.1029/2007JD009064.

K.M. Markowicz et al. / Atmospheric Environment 48 (2012) 66e75 Ellrod, G.P., Connell, B.H., Hillger, D.W., 2003. Improved detection of airborne volcanic ash using multispectral infrared satellite data. J. Geophys. Res. 108 (D12), 4356. doi:10.1029/2002JD002802. EUMETSAT, 2007. Volcanic Ash Detection with MSG Algorithm Theoretical Basics Documents EUM/MET/REP/07/0467. European Commission: Volcano crisis report, 2010. http://ec.europa.eu/transport/ash_cloud_crisis_en.htm (last access: November 2010). European Commission, 2010. Volcano Crisis Report: Report on the actions undertaken in the context of the impact of the volcanic ash cloud crisis on the air transport industry. http://ec.europa.eu/transport/doc/ash-cloud-crisis/2010_ 06_30_volcano-crisis-report.pdf (last access: 30.06.2010). Fernald, F.G., 1984. Analysis of atmospheric lidar observations: some comments. Appl. Opt. 23, 652e653. Flentje, H., Heese, B., Reichardt, J., Thomas, W., 2010. Aerosol profiling using the ceilometer network of the German Meteorological Service. Atmos. Meas. Tech. Discuss. 3, 3643e3673. doi:10.5194/amtd-3-3643-2010. Frey, S., Poenitz, K., Teschke, G., Wille, H., 2010. Detection of Aerosol Layers with Ceilometer and the Recognition of the Mixed Layer Depth. ISARS. pp. 3646e3647. Gasteiger, J., Groß, S., Freudenthaler, V., Wiegner, M., 2010. Volcanic ash from Iceland over Munich: mass concentration retrieved from ground-based remote sensing measurements. Atmos. Chem. Phys. Discuss. 10, 26705e26750. doi:10.5194/acpd-10-26705-2010. 2010.5562. Hansen, J., Lacis, A., Ruedy, R., Sato, M., 1992. Potential climate impact of Mount Pinatubo eruption. Geophys. Res. Lett. 19 (2), 215e218. doi:10.1029/ 91GL02788. Harrison, L., Michalsky, J., Berndt, J., 1994. Automated multifilter rotating shadowband radiometer: an instrument for optical depth and radiation measurements. Appl. Opt. 33, 5118e5125. Harshvardhan, 1979. Perturbation of the zonal radiation balance by stratospheric aerosol layer. J. Atmos. Sci. 37, 1274e1285. Heese, B., Flentje, H., Althausen, D., Ansmann, A., Frey, S., 2010. Ceilometerelidar inter-comparision: backscatter coefficient retrieval and signal-to-noise ratio determination. Atmos. Chem. Phys. Discuss. 3, 3907e3924. doi:10.5194/amtd-33907-2010. Hess, M., Koepke, P., Schult, I., 1998. Optical properties of aerosols and clouds: the software package OPAC. Bull. Am. Met. Soc. 79, 831e844. IATA: Press Release, 2010. Volcano Crisis Cost Airlines $1.7 Billion in Revenue e IATA Urges Measures to Mitigate Impact. http://www.iata.org/pressroom/pr/Pages/ 2010-04-21-01.aspx (last accessed 2.02.11). Kirchner, I., Stenchikov, G., Graf, H.-F., Robock, A., Antuna, J., 1999. Climate model simulation of winter warming and summer cooling following the 1991 Mount Pinatubo volcanic eruption. J. Geophys. Res. 104, 19,039e19,055. Klett, J.D., 1985. Lidar inversions with variable backscatter/extinction values. Appl. Opt. 24, 211e220. Li, Z., Goloub, P., Blarel, L., Damiri, B., Podvin, T., Jankowiak, I., 2007. Dust optical properties retrieved from ground-based polarimetric measurements. Appl. Opt. 46, 1548e1553. Lindqvist, H., Nousiainen, T., Zubko, E., Muñz, O., 2011. Optical modeling of vesicular volcanic ash particles. J. Quant. Spectrosc. Radiat. Transfer 112, 1871e1880. Maciszewska, A., Markowicz, K., Witek, M., 2010. Multiyear analysis of the aerosol optical thickness over Europe. Acta Geophys. 58, 1147e1163. Markowicz, K.M., Flatau, P.J., Kardas, A.E., Stelmaszczyk, K., Woeste, L., 2008. Ceilometer retrieval of the boundary layer vertical aerosol extinction structure. J. Atmos. Ocean. Techn. 25, 928e944. Mather, T.A., Pyle, D.M., Oppenheimer, C., 2003. Tropospheric volcanic aerosol. In: Robock, A., Oppenheimer, C. (Eds.), Volcanism and the Earth’s Atmosphere, Geophysical Monograph, vol. 139. Am. Geophys. Union, Washington, D.C, pp. 189e212.

75

McCormick, M.P., Thomason, L.W., Trepte, C.R., 1995. Atmospheric effects of the Mt Pinatubo eruption. Nature 373, 399e404. doi:10.1038/373399a0. Mona, L., Amodeo, A., Boselli, A., Cornacchia, C., D’Amico, G., Giunta, A., Madonna, F., Pappalardo, G., 2010. Observations of the Eyjafjallajökull eruption’s plume at Potenza EARLINET station. Geophys. Res. Abs 12 EGU2010e15,747. Myhre, G., Myhre, A., Stordal, F., 2001. Historical evolution of radiative forcing of climate. Atmos. Environ. 35, 2361e2373. O’Connor, E.J., Illingworth, A.J., Hogan, R.J., 2004. A technique for auto-calibration of cloud lidar. J. Atmos. Oceanic Technol. 21, 777e786. Papparlardo, G., Mattis, I., EARLINET Team, 2010. Dispersion and Evolution of the Eyjafjallajökull Ash Plume over Europe: Vertically Resolved Measurements with the European LIDAR Network EARLINET, vol. 12. EGU General Assembly, Vienna. EGU2010e15731. Parker, D.E., Wilson, H., Jones, P.D., Christy, J.R., Folland, C.K., 1996. The impact of Mount Pinatubo on worldwide temperatures. Int. J. Climatol. 16, 487e497. Pavolonis, M.J., Feltz, W.F., Heidinger, A.K., Gallina, G.M., 2006. A daytime complement to the reverse absorption technique for improved automated detection of volcanic ash. J. Atmos. Ocean. Techn. 23, 1422e1443. Pietruczuk, A., Podgórski, J., 2009. The lidar ratio derived from sun-photometer measurements at Belsk Geophysical Observatory. Acta Geophys. 57 (2), 476e493. Pietruczuk, A., Krzyscin, J.W., Jaroslawski, J., Podgorski, J., Sobolewski, P., Wink, J., 2010. Eyjafjallajökull volcano ash observed over Belsk (52 N, 21 E), Poland, in April 2010. Int. J. Remote Sensing 31, 3981e3986. Porter, J.N., Lienert, B., Sharma, S.K., 2000. Using the horizontal and slant lidar calibration methods to obtain aerosol scattering coefficients from a coastal lidar in Hawaii. J. Atmos. Oceanic Technol. 17, 1445e1454. Porter, J.N., Horton, K., Mouginis-Mark, P., Lienert, B., Lau, E., Sharma, S.K., Sutton, J., Elias, T., Oppenheimer, C., 2002. Lidar and sun photometer measurements of the Hawaii Pu’u O’o volcano plume: estimates of SO2 and aerosol flux rates and SO2 Lifetimes. Geo. Res. Let. 29 2002GL014744. Sasano, Y., Browell, E.V., Ismail, S., 1985. Error caused by using a constant extinction/ backscattering ratio in the lidar solution. Appl. Opt. 24, 3929e3932. Stachlewska, I.S., Markowicz, K.M., 2010. On forward Klett’s inversion of ceilometer signals, 25th ILRC International Laser Radar Conference, 5e9 July 2010, St. Petersburg, Russia. Stachlewska, I.S., Ritter, C., 2010. On retrieval of lidar extinction profiles using TwoStream and Raman techniques. Atmos. Chem. Phys. 10, 2813e2824. doi:10.5194/ acp-10-2813-2010. Stohl, A., Hittenberger, M., Wottowa, G., 1998. Validation of the Lagrangian particle dispersion model FLEXPART against large scale tracer experiments. Appl. Opt. 32 (24), 4245e4263. doi:1016/S1352-2310(98)00184-8. Stohl, A., Prata, A.J., Eckhardt, S., Clarisse, L., Durant, A., Henne, S., Kristiansen, N.I., Minikin, A., Schumann, U., Seibert, P., Stebel, K., Thomas, H.E., Thorsteinsson, T., Tørseth, K., Weinzierl, B., 2011. Determination of time- and height-resolved volcanic ash emissions for quantitative ash dispersion modeling: the 2010 Eyjafjallajökull eruption. Atmos. Chem. Phys. Discuss. 11, 5541e5588. doi:10.5194/acpd-11-5541-2011. Thomas, H.E., Prata, A.J., 2010. Sulphur dioxide as a volcanic ash proxy during the April/May 2010 eruption of Eyjafjallajökull Volcano, Iceland. Atmos. Chem. Phys. Discuss. 11, 7757e7780. Trigo, R.M., Vaquero, J.M., Alcoforado, M.J., Barriendos, M., Taborada, J., CarciaHerrera, R., Luterbacher, J., 2009. Iberia in 1816, the year without a summer. Int. J. Clim. 29, 99e115. doi:10.1002/joc.1693. Wang, X., Boselli, A., D’Avino, L., Pisani, G., Spinelli, N., Amodeo, A., Chaikovsky, A., Wiegner, M., Nickovic, S., Papayannis, A., Perrone, M.R., Rizi, V., Sauvage, L., Stohl, A., 2008. Volcanic dust characterization by EARLINET during Etna’s eruptions in 2001e2002. Atmos. Environ. 42 (5), 893e905. Welton, E.J., Campbell, J.R., 2002. Micropulse lidar signals: uncertainty analysis. J. Atmos. Oceanic Technol. 19, 2089e2094.