ÓPTICA PURA Y APLICADA – Vol. 37, núm. 3 - 2004
Aerosol optical thickness and its spectral dependence derived from sun
photometer measurements over the southern North Sea coastal region
I. Behnert, V. Matthias, R. Doerffer
Institute for Coastal Research, GKSS Research Center, Max-Planck-Straße 1,
Geesthacht D-21502, Germany, [email protected]
REFERENCES AND WEB LINKS. [1] http://envisat.esa.int/instruments/meris/
[2] Y.J. Kaufman, D.Tanre, H.R. Gordon, T. Nakajima, J. Lenoble, R. Frouin, H. Grassl, B.M. Herman, King M.D. and P.M. Teillet, “Passive remote sensing of tropospheric aerosol and atmospheric correction for the aerosol effect”, J. Geophys. Res., 102/D14, 16815-16830, (1997).
[3] H. R. Gordon, “Atmospheric correction of ocean color imagery in the Earth Observing System era”, J. Geopys. Res., 102/D14,17081-17106 (1997).
[4] WCP–112: World Climate Research Program – “A preliminary cloudless standard atmosphere for radiation computation”. Int. Ass. For Meteor. and Atm. Physic., Radiation Commission, WMO/TD-NO.24., 53 pp. (1986).
[5] E.P. Shettle and R.W. Fenn, “Models for the aerosols of the lower atmosphere and the effects of humidity variations on their optical properties”, AFCRL Tech, Rep. 790214, Air Force Cambridge Research Laboratory, Hanscom Air Force Base, MA, 100 pp. (1979).
[6] P. Koepke, M. Hess, I. Schult and E. P. Shettle, “Global aerosols data set”. MPI Metorologie Hamburg Rep. 243, 44 pp. (1997).
[7] M. Hess, P. Koepke and I. Schult, “Optical properties of aerosols and clouds: The software package OPAC”, Bull. Amer. Meteor. Soc., 79, 831-844 (1998).
[8] http://aeronet.gsfc.nasa.gov/
ABSTRACT:
The aerosol optical thickness and its spectral dependence for the southern North
Sea region were derived from ground-based monitoring using sun photometer
AERONET data of Helgoland, Hamburg, Oostende and Lille. Their annual
variations were investigated based on 2-4 years of measurements. Differences
and similarities among the sites can be explained by the different air mass origins
as well as by different local sources. This regional information on aerosols will
be used to improve the atmospheric correction algorithm for MERIS data
(Medium Resolution Imaging Spectrometer, ENVISAT), when captured over
coastal waters.
[9] B.N. Holben, T.F.Eck, I.Slutsker, D.Tanre, J.P.Buis, A.Setzer, E.Vermote, J.A.Reagan, Y.Kaufman, T.Nakajima, F.Lavenu, I.Jankowiak, and A.Smirnov, “AERONET - A federated instrument network and data archive for aerosol characterization”, Rem. Sens. Environ., 66, 1-16, (1998).
[10] E.M., Rollin, “An introduction to the use of Sun-photometry for the atmospheric correction of airborne sensor data”. Activities of the NERC EPFS in support of the NERC ARSF. ARSF Annual Meeting, Keyworth, Nottingham, UK, (2000).
[11] O. Dubovik, and M. D. King, “A flexible inversion algorithm for retrieval of aerosol optical properties from Sun and sky radiance measurements," J. Geophys. Res., 105, 20 673-20 696, (2000).
[12] O. Dubovik, A. Smirnov, B. N. Holben, M. D. King, Y.J. Kaufman, T. F. Eck, and I. Slutsker, “Accuracy assessments of aerosol optical properties retrieved from AERONET sun and sky-radiance measurements”, J. Geophys. Res., 105, 9791-9806, (2000).
[13] V. Matthias, D.Balis, J. Bosenberg, R.Eixmann, M.Iarlori, L.Komguem, I. Mattis, et al., “The vertical aerosol distribution over Europe: Statistical analysis of Raman lidar data from 10 years EARLINET stations”, J.Geoph.Res., (accepted), (2004).
[14] B.N. Holben, D.Tanre, A.Smirnov, T.F.Eck, I.Slutsker, N.Abuhassan, W.W.Newcomb, J.Schafer, B.Chatenet, F.Lavenue, Y.J.Kaufman, J.Vande Castle, A.Setzer, B.Markham, D.Clark, R.Frouin, R.Halthore, A.Karnieli, N.T.O'Neill, C.Pietras, R.T.Pinker, K.Voss, and G.Zibordi, “An emerging ground-based aerosol climatology: Aerosol Optical Depth from AERONET”, J. Geophys. Res., 106, 12067-12097, (2001).
[15] Y.J. Kaufman, A.Smirnov, B.N.Holben, and O.Dubovik, “Baseline maritime aerosol: methodology to derive the optical thickness and scattering properties”, Geoph.Res.Lett., 28, 3251-3254, (2001).
[16] N.T., O'Neill, A. Ignatov, B.N., Holben and T.F., Eck, “The lognormal distribution as a reference for reporting aerosol optical depth statistics; empirical tests using multi-year, multi-site AERONET sunphotometer data”, Geophy. Res. Lett., Vol. 27, pp. 3333-3336, (2000).
[17] V. Matthias and J. Boesenberg, “Aerosol climatology for the planetary boundary layer derived from regular lidar measurements”, Atmospheric Research, 63, 221-245, (2002).
[18] A. Smirnov, B.N.Holben, Y.J., Kaufman, O. Dubovik, et. al., “Optical properties of atmospheric aerosols in maritime environments”, J. Atm.Sci., 59, pp. 501-523, (2002).
[19] O. Dubovik, B.N.Holben, T.F.Eck, A.Smirnov, Y.J.Kaufman, M.D.King, D.Tanre, and I.Slutsker, “Variability of absorption and optical properties of key aerosol types observed in worldwide locations”, J.Atm.Sci., 59, 590-608, (2002).
1.- Introduction
The application of MERIS data (Medium Resolution Imaging Spectrometer on ENVISAT started in February 2002) [1] for remote sensing of turbid coastal waters requires a special atmospheric correction. It has to take the backscattering by water constituents in the near infrared spectral range into account but also aerosols with optical properties which differ from those present over open oceans. Regions for our investigation are the coastal areas of the southern North Sea, where the atmosphere is characterised by a strong anthropogenic influence. Higher aerosol loading and urban absorbing aerosols are expected to be specific for this region. This requires the use of an atmospheric algorithm which takes multiple scattering into account [2],[3].
We expect that the standard aerosol models [4],[5] or the aerosol optical properties calculated for the average maritime, urban and continental conditions [6],[7] do not cover the whole range of the aerosol types and their related concentrations of this region, as required for realistic atmospheric
radiative transfer calculations and, thus, the design of the atmospheric correction procedure.
The present study is focused on the spatial and the temporal variation of the aerosol optical thickness over the southern North Sea region, derived from sun photometer measurements over 4 years in the framework of the AERONET [8]. The spectral dependence of the aerosol optical thickness represented by the Angstrom coefficient is studied with the goal to identify different mixtures of aerosol basic components (maritime, urban-industrial, dust, biomass burning) for each site.
The comparison of the data from various sites, located in an area with a diameter of about 1500km, could be used to relate differences in aerosol optical properties to different air mass origins and local aerosol sources.
2.- Methods and Data
2.a.- Method
The aerosol optical thickness τa(λ) and the
Angstrom coefficient α can be derived with relatively high accuracy from the extinction measurements of the direct solar beam (sun-radiance measurements). They are also the key parameters for the validation of the atmospheric correction of remote sensing data.
The sun and sky-radiance information is used to derive the spectral single scattering albedo, the refractive index and the aerosol size distribution [11], [12]. These parameters are calculated with lower accuracy than the τa(λ) and their retrieval is limited to certain conditions (sun zenith angle
θs≥45°, cloudless and homogeneous sky with
symmetric almucantars at 21 angles, τa(440) ≥0.4). The number of the available almucantar measurements, which fulfill these conditions, is relatively low at the northern European sites. Thus, the related data can not be used for statistic purposes as it is the case with the optical thickness data.
2.b.- Data
The τa(λ) is determined at 7 wavelengths
(340, 380, 440, 500, 670, 870, 1020 nm). The Angstrom coefficient α440−870 is computed for the
visible spectral region. The AERONET sites selected for the present study are represented in the Figure.1. Two maritime sites, supposed to have less antropogenic influence, were additionally included (Rame Head, Azores Island) for comparisons.
Figure.1. AERONET sites from the southern North Sea region and maritime sites (Azores, Rame Head)
Description of the sites:
(1) Helgoland Island (N 54°10’, E 07°53’): rocky island, area is 1 km2, situated at 70 km from land.
(2) Hamburg (N53°34', E 09°58'): town of 1.8 million habitants, situated in a flat terrain at about 100 km from the coast.
(3) Oostende (N 51°13', E 02°55'): highly populated Belgian port city.
(4) Lille (N 50°36', E 03°08'): highly populated town at about 70 km from the coast.
(5) Rame Head (N 50°21', W 04°08'): town located at the English Channel in a low populated area. (6) Azores Island (N 38°31', W 28°37'): located in the Atlantic Ocean at about 5000 km west off the Portuguese coast.
3.- Results
Annual variation of τa(500)
Figure.2. Annual variation of the τa(500) using daily
means and a gliding average over 31 days.
to the springtime (April–May) and (2) to the summertime (July-August).This annual variation is not visible at the maritime sites (Rame Head and Azores).
A similar annual variation was determined from regular lidar extinction profiles over different German sites between 2000-2002 in the framework of the EARLINET project [13].
The summertime maximum is explained by Holben et al. [14] by the accumulation of high aerosol concentrations over the midlatitude regions, because the air mass circulation slows down at this time and smog is produced. An argument in favour of of the smog formation is that the highest values of
α440−870 monthly average are reached also during the
July-August period for all sites (Table I).
The spring peak is mentioned at other Atlantic sites without explanation [15]. This aerosol
optical thickness peak is not reflected in the
α440−870 monthly average values (Table I). A possible
explanation is that higher levels of relative humidity produce a growth of the aerosol particles, known to be most effective for the larger particles [5]. The spectral extinction slope decreases with the increase of the relative humidity, and thus, the aerosol size. That means that an increase of α440−870cannot be
observed, even if τa(500) is increasing at the same
time.
Further investigations of the relative humidity as well as of the air mass back-trajectories for this period over the North Europe region are required.
TABLE I
Annual variation of the Angstrom coefficient, α440−870 (monthly average) and σ (standard deviation) for N days
The maximum of the α440−870 monthly averages is written with bold, generally related to July or August months
Month Helgoland α
440−870 /σ
Hamburg
α440−870/σ
Oostende
α440−870/σ
Lille
α440−870/σ
Rame Head
α440−870/σ
Azores
α440−870/σ
Jan - 1.39/0.41 - 1.26/0.61 0.51/0.18 0.26/0.13
Feb - 1.43/0.32 - 1.16/0.30 0.83/0.39 0.51/0.32
Mar - 1.43/0.33 1.26/0.31 1.13/0.44 - 0.44/0.26
Apr 1.24/0.39 1.22/0.42 1.30/0.45 1.18/0.45 1.01/0.45 0.61/0.24 May 1.11/0.35 1.43/0.37 1.18/0.43 1.25/0.37 1.18/0.28 0.75/0.31 Jun 1.49/0.42 1.46/0.30 1.59/0.34 1.49/0.32 0.93/0.35 1.02/0.35 Jul 1.71/0.36 1.69/0.20 1.61/0.39 1.54/0.42 1.28/0.44 1.15/0.46 Aug 1.30/0.43 1.51/0.35 1.51/0.22 1.47/0.30 1.46/0.31 1.18/0.35 Sep 1.29/0.45 1.31/0.44 1.39/0.29 1.29/0.38 0.98/0.45 0.95/0.47 Oct 1.27/0.50 1.51/0.41 0.82/0.39 1.02/0.42 1.25/0.52 0.74/0.37 Nov - 1.37/0.48 0.87/0.54 1.05/0.40 0.68/0.50 0.57/0.37 Dec - 1.43/0.12 0.72/0.66 1.10/0.45 0.95/0.98 0.33/0.26
N 170 286 296 413 119 405
Classification of the AERONET sites
The frequency distribution of the τa(λ)
follows a lognormal distribution [16], [17]. This distribution is the same for all measurement points or daily mean values. The daily means instead of all points are considered for computing the statistics, in order not to overestimate the influence of the long summer days data. The characteristics of the distribution for the τa(500) and the α440−870 daily
means are reported for the sites and compared with those from the maritime ones (Table II).
The summer time series (April-September) of 2000(2001)-2003 is considered for the further analysis, because this was the period when all sites
were in operation, with the exception of the Rame Head, with data only from 1997-1998.
Lille, Hamburg and Oostende with τa(500) median values between 0.21 and 0.23 are urban sites from northern Europe. Helgoland with 0.194 shows only 10% lower values and can be classified as an urban-maritime site. Rame Head with 0.134 is a maritime site with some continental influence. Azores with median values about 0.1 is similar to other Atlantic maritime sites as Bermuda and Ascension Islands [18].
The α440-870 values for all the sites with the
such as soot. The median value of α440-870 of the
Azores is, as expected, similar to the mode value determined for Bermuda and Ascension (Atlantic Ocean) [18], demonstrating the low anthropogenic influence.
TABLE II
Median and standard deviation (on log scale) for τa(500)
and α440-870 daily means, N= number of days
AERONET
sites N 500nm med/sigma
α440-870
med/5%/95% Helgoland 163 0.194/0.598 1.60/0.61/2.08
Hamburg 204 0.209/0.607 1.66/0.67/2.02 Oostende 250 0.225/0.585 1.60/0.79/1.96
Lille 322 0.229/0.553 1.54/0.73/1.87 Rame Head 88 0.134/0.552 1.57/0.58/1.93
Azores 237 0.099/0.513 0.79/0.28/1.60
Figure.3. represents the variation of α440-870
with τa(500) for all sites. Similarities between the
coastal sites (Helgoland, Oostende) and the two northern European town sites (Hamburg, Lille) are found. The highest values of α440-870 are related to the highest values of τa(500). This pattern
demonstrates again the dominance of the urban aerosols at all these sites.
Azores and Rame Head are characterised by a relatively low aerosol optical depth. The high values of α440-870 for low aerosol loadings could be due to the error in the calculation of the α440-870 from
very low values of the aerosol optical thickness.
Figure.3. Variation of the Angstrom coefficient with the aerosol optical thickness at 500nm (black: coastal sites; red: town sites; blue: maritime sites).
Correlation analysis between Helgoland and the other coastal sites
Days with simultaneous measurements at Helgoland and the other AERONET sites demonstrate that Helgoland data is well correlated with that from Hamburg (Table III). The correlation factor between Helgoland and Hamburg is about 0.6, for daily means of common measurement days. The correlation decreases with increasing distance between stations. The correlation coefficient between Helgoland and Azores is about 0.
TABLE III
Correlation of the τa(500) daily means between the sites,
considering common measurements days from 2000-2003
AERONET sites commondays coef. 500nm correlation Helgoland-Hamburg 101 0.5888 Helgoland-Oostende 95 0.3986
Helgoland-Lille 153 0.2000 Helgoland-Azores 75 -0.0651
Conclusions.
The comparison of the most basic aerosol optical properties (aerosol optical thickness and its spectral dependence) demonstrated that obvious similarities exist between the northern European AERONET sites. The high similarities are related to: (1) the annual variation of the aerosol loading with two maximums in spring and the summer (2) all sites with the exception of Azores are characterized by α440-870 median values greater than 1.5 which are
related to the presence of urban aerosols.
The present work will be continued with a comparative study of the radiative properties as: the spectral single scattering albedo, refractive index and the volume size distribution, even if very few retrieved data over 4 years are available. The retrieved values for the northern European sites will be compared with those of the “key aerosol types” [19].
Acknowledgments.
We thank the AERONET Principal Investigators and their staff for establishing and maintaining the Azores, Hamburg, Lille, Oostende and Rame Head sites used in this investigation.
Thanks to Peter Kipp and Firma Elektronik Jebo for establishing and maintaining the Helgoland AERONET site.