- Email: [email protected]
Aerosol and cloud characteristics analysis methods using multiple kinds of Raman lidar signals
Optik 124 (2013) 1170–1174
Contents lists available at SciVerse ScienceDirect
Optik journal homepage: www.elsevier.de/ijleo
Aerosol and cloud characteristics analysis methods using multiple kinds of Raman lidar signals Imkang Song a , Il-Moon Chae a , Sung-Hoon Baik b , Sunngchul Choi b , Dukhyeon Kim c,∗ , Sunho Park c , Yonggi Kim d a
EngiOn Co., Ltd, 1-508, INNOPLEX, 552-2 Woncheon-dong, Yeongtong-gu, Suwon, 443-380, Republic of Korea Laboratory for Quantum Optics, Korea Atomic Energy Research Institute, Daejeon 305-353 Republic of Korea c Division of Cultural Studies, Hanbat National University, Daejeon 305-719, Republic of Korea d Department of Physics, Kongju National University, Kongju 314-701, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 15 October 2011 Accepted 4 March 2012
Keywords: Rotational Raman lidar Color Ratio Cloud Aerosol Liquid water
a b s t r a c t In this paper, we introduce three kinds of methods for the analysis of aerosol and cloud droplet characteristics: backscattering color ratio at two wavelengths (color ratio), aerosol liquid-water content, and cloud droplet size distribution using the liquid water and aerosol extinction coefficients. Based on theoretical perspective as well as our experimental results, we find that the liquid-water Raman scattering efficiency does not depend on four orders of particle size, but rather, depends on less than four orders, particularly in a smaller effective size distribution than Mie scattering efficiency. Therefore, we conclude that the color ratio method can be applied to aerosols, while the ratio between the liquid water Raman and extinction coefficients can be applied to cloud droplet size measurements. © 2012 Elsevier GmbH. All rights reserved.
1. Introduction Aerosol characteristics are an important parameter in long-term climate, and short-term weather changes because the aerosol scattering phase function and the physics of cloud formation depend upon aerosol size, shape, and aerosol chemical-optical characteristics. Also, a variety of natural and anthropogenic aerosol sources not only endow clouds with more reflectivity, but also lengthens the lifetime of clouds by suppressing precipitation [1]. However aerosol and liquid water parameters are known only on a local basis with low time resolutions and a limited range of characteristics, because most scientists have used traditional point-measuring equipments. Many researchers have tried to measure a variety of aerosol chemical and physical properties such as chemical components, physical characteristics, size distribution, and the refractive index by using a multi-wavelength Raman lidar and multiple-wavelength depolarization lidar. Tartarov et al. [2] provided a number of signatures of the chemical components of atmospheric aerosols using a multi-channel lidar spectrometer, but the spectrometer they used did not have a sufficient rough spectral resolution or spectral range. They also required too much time because of the low aerosol Raman scattering cross-section. Shimizu et al. [3] observed Asian dust and other aerosol characteristics using a polarization lidar. However,
∗ Corresponding author. E-mail address: [email protected] (D. Kim). 0030-4026/$ – see front matter © 2012 Elsevier GmbH. All rights reserved. doi:10.1016/j.ijleo.2012.03.012
the depolarization method is an indirect method, and therefore, has a complex dependence upon aerosol shape and aerosol alignment in the atmosphere. Thus, this method has a limited sphere of application. Furthermore, dust is an aerosol type that can be misclassified as a cloud when the dust layer is very dense [4]. Also Asian dust events are normally transported by cold fronts, and the transported dust is frequently adjacent to, embedded in, or mixed with clouds [5]. Satellite measuring systems can give global and 24 h information about aerosol extinction or backscattering coefficients, but they have number of limitations as regards spatial and time resolutions. Because these satellite data have limitations in obtaining certain important information about aerosol chemical or physical parameters and their altitude distributions, we need to find another method that can calibrate satellite data. From meteorological and health point of view, both the aerosol parameters and spatial distribution are important, so we must measure aerosol characteristics for a large spatial distribution as well. The concentration and profile of liquid water and water vapor can be measured by using a radiometer and a GNSS system with a low spatial resolution [6], but aerosol hygroscopic parameters cannot be measured via these methods. Cloud droplet size distribution depends upon the aerosol particle distribution and aerosol hygroscopic characteristics. In addition, aerosol atmospheric liquid water is composed of water droplets that are attached to the aerosol nucleus. To specify aerosol’s hygroscopic characteristics, we must also measure these characteristics.
I. Song et al. / Optik 124 (2013) 1170–1174
1171
Recently, other scientists have tried to combine more than two parameters in order to distinguish among aerosols by using approaches such as the lidar ratio, depolarization, and color ratio etc. [7,8]. If we can simultaneously measure a liquid water lidar signal, a nitrogen Raman lidar signal, and an aerosol Mie lidar signal, we can characterize an aerosol’s hygroscopic characteristics and its cloud droplet size. The main purpose of this study is to introduce such a method of measurement and to demonstrate it in terms of aerosol hygroscopic characteristics, cloud droplet size, and particle size, using liquid water Raman, rotational, and Mie lidar systems. 2. Method Fig. 1. Mie and liquid water Raman scattering efficient [10].
Although many scientists have tried to measure aerosols by using various lidar systems, their approaches have all treated water cloud droplets and aerosols simultaneously. However, cloud droplets have a constant refractive index and a specific size distribution that depend upon cloud height and other conditions. Here, we discuss aerosol hygroscopic characteristics, aerosol effective size distribution, and cloud droplet size distribution from a number of theoretical viewpoints.
smaller than the water droplet Raman efficiency, but as the droplet size increases, the Mie scattering efficiency increases more quickly than the Raman scattering. This means that when the liquid water lidar signal is normalized by an aerosol lidar signal, the normalized liquid water Raman values are dependent upon the aerosol droplet size.
2.1. Method for obtaining cloud droplet size distribution
2.2. Methods for obtaining aerosol size distributions
Mie scattering lidar signals contain information about the aerosol backscattering and extinction coefficients, while liquid water Raman lidar signals contain information about the total liquid water volume that is attached to the aerosol. Whiteman and Melfi [9] have suggested a simple method for measuring water droplet size. They were able to measure cloud droplet size using liquid water Raman lidar signals and extinction coefficients. As calculated by Veselovskii et al. [10], the Raman scattering efficiency of liquid water is constant (QLiquid Raman ≈ 4), and is approximately twice the extinction efficiency (Qext ≈ 2) when the size parameter x (2r/) is greater than 1. Hence, the total Raman signal of liquid water depends only upon the total volume of liquid water, as follows:
The backscatter ratio at different frequencies–the so called “color ratio”–depends upon the size distribution, the source of the aerosol (refractive index), and the shape of the aerosol particles, and is particularly useful for inferring the mean size. Ansmann et al. [12] suggest how the color ratio, P and S may be used to infer aerosol properties, and they offer validation of their approach as obtained from aircraft flights. Tackett et al. [13] also provide the first lidar observations of the backscatter-color ratio, and show that these quantities are directly related to aerosol properties. They show that the backscatter and color ratio are enhanced adjacent to the cloud edge, particularly near the cloud top and cloud base. Specifically, there is an increase of 31 ± 3% and 42 ± 2% in the layer-integrated median backscatter at wavelengths of 532 nm and 1064 nm, respectively, and the layer-averaged color ratio increases by 15 ± 5%. These backscatter calculations suggest that our observations mode adjacent to clouds are best explained by the aerosol size distribution with a reduced number concentration, increased median radius, and decreased width, as compared to observations made at a distance from clouds. B. de Foy et al. [14] also suggest four intensive parameters that depend only on the aerosol type and not on the concentration that can be calculated from the measurements: the lidar ratio, the aerosol depolarization ratios at 532 nm and 1064 nm, the backscatter coefficient, and the color ratio. Recently, Burton et al. [15] and Sugimoto et al. [16] suggested that although the depolarization color ratio has not been studied extensively, it can provide information about the relative size of non-spherical particles and may also provide a means of distinguishing between smoke and pollution.
cloud ˇLiquid
Raman
cloud n(r)QLiquid
=
≈
n(r)
Raman
(r)
4r 3 dr 3
16r 3 dr. 3
(1)
At the same time, the extinction coefficient depends upon the total area, as follows:
∞
n(r)Qext r 2 dr ≈
˛= 0
∞
n(r)2r 2 dr
(2)
0
Because we can easily measure a cloud’s optical and physical depth from the bottom of the cloud, we can easily determine a cloud’s extinction coefficient [7]. Consequently, from Eqs. (1) and (2), the effective size of the water droplet can be expressed as Eq. (3).
∞ cloud 3ˇLiquid /16 n(r)r 3 dr Raman reff = 0∞ = . 0
n(r)r 2 dr
˛/2
2.3. Method for obtaining aerosol hygroscopic characteristics (3)
cloud where ˇLiquid and ˛ are the water droplet cloud backscattering Raman coefficient and the extinction coefficient, respectively. These coefficients can be obtained using conventional lidar signal analysis methods [11]. Fig. 1 shows the liquid water Raman and Mie scattering efficiency of a water droplet [10]. As can be seen in the figure, when the aerosol droplet size is small, the Mie scattering efficiency is
Normally, atmospheric aerosol contains hydrogen-bonded water, a kind of water that has a different spectrum from aerosols that contain water vapor or bulk water. If we simultaneously measure the Raman signal of this bonded water and the aerosolscattering Mie signal, we can then extract useful information about the hygroscopic characteristics of the aerosol. For example, in a non-hygroscopic aerosol we find a strong Mie scattering signal even though we find a weak liquid water Raman lidar signal, and vice versa. When the Raman and Mie lidar receivers have the same lidar
1172
I. Song et al. / Optik 124 (2013) 1170–1174
Fig. 2. Schematic diagram of experimental system.
overlap function and measure the same direction, we can write the two lidar equations as follows:
3. Experimental setup and results
the rotational Raman signal, the nitrogen vibrational Raman signal for normalization, and a full spectrum (32 channels) of the vibrational Raman signal for water vapor/liquid water. This system can block out an elastic signal (355 nm) of more than 14 orders of magnitude and can resolve Raman wavelengths from 401 nm to 410 nm using 32-channel PMT sensors (Hamamatsu H7260A) [17,18]. From the 355 nm rotational Raman and 532 nm rotational Raman and elastic Mie signals, we can calculate the aerosol backscattering coefficient without any assumptions [19]. Fig. 3 shows the traditional elastic, rotational Raman, and the nitrogen rotational Raman lidar signals obtained at these two wavelengths. As can be seen in the figure, a strong elastic backscatter signal was measured in the presence of clouds at around 4.5 km, while the rotational Raman signal disappeared in this region. There was no leakage of the elastic signal into the Raman channel, even in densely cloudy conditions. Fig. 4(a) and (b) shows the changes in the of elastic, color ratio, and liquid water Raman/extinction ratio at different altitudes in the boundary layer near a cloud on November 14, 2007. The color ratio and liquid water Raman/extinction coefficient ratio can give information about the change in particle size. In the figures, it can be seen that even though the boundary layer is at the same altitude (1800 m), its range changes with time. As regards to the cloud conditions, the cloud’s bottom height was 2100 m at 20:35, but after the boundary layer collapsed, the bottom height became lower at 22:50 (1900 m) than it was at 20:35. In this case, the three parameters show very similar patterns in the mist (1400–1600 m), the boundary layer (1800 m), and the cloud (2200–2400 m). From these figures, we can deduce that the aerosol concentration and aerosol size distribution change. Fig. 5 shows the change in the elastic, color ratio, and liquid water Raman/extinction coefficient ratio near the cloud on August 20, 2007 (following a rainy day). This figure shows the changes in the elastic, color ratio, and liquid water Raman/extinction coefficient ratio near the cloud at different altitudes. The color
Fig. 2 shows a schematic diagram of the installed lidar system. A pulsed Nd:YAG laser was used as a transmitter and produced 6 ns pulses at a repetition rate of 30 Hz, which a maximum energy of 150 mJ/pulse at 355 nm, and 300 mJ at 532 nm. To decrease the degree of laser beam divergence, a beam expander was used in the transmitting system that could expand a laser beam 10 times. The two wavelength (355 nm, 532 nm) lidar signals obtained from two telescopes were the input to the fiber-based double-grating monochrometer. From the hardware perspective, the receiving system has three conceptual configurations. The first conceptual channel is water vapor, liquid water and ice water channels, and the second conceptual channel consists of rotational Raman channels for 355 nm and 532 nm. The third configuration is a Mie scattering channel. From the 355-nm wavelength laser, we can receive rotational Raman, elastic Mie, and water Raman signals. The receiving system consisted of four signals from as many detectors: the elastic signal,
Fig. 3. Traditional return signals for elastic, rotational Raman, and nitrogen rotational Raman LIDAR signals on October 12, 2007 (KST). The elastic signal was not affected in the rotational Raman channel.
PMie,Raman (z) = K(z)
ˇMie,Raman z2
e
−
z 0
˛(z ,0 )+˛(z ,Mie,Raman ) dz
.
(4)
Here, K, , ˇ, z, ˛, denote a constant, the geometric overlap function, the backscattering coefficient, distance (z), the extinction coefficient, and the wavelength for a given subscript, such as liquid water Raman (Liquid Raman, Vapor Raman), nitrogen molecule, and aerosol (Mie), and other Raman (R) wavelengths. When the liquid-water lidar signal is divided by the aerosol and water vapor Raman signals, the ratios are approximately written as follows: R1 (z) =
R2 (z) =
PLiquid Pvapor (z) PLiquid PMie (z)
,
=
(5) ˇRaman e ˇMie
z 0
˛(z ,Raman )−˛(z ,Mie ) dz
.
(6)
From R1 of Eq. (5) we can obtain some relative information about a liquid water cloud. An increase in the value (R1 ) time indicates an increase in the cloud droplet concentration or a decrease in the cloud water vapor. Therefore this value indicates the relative liquid water content of a given water vapor. Changes in this R1 indicate changes in the phase of water from vapor to liquid. If this value decreases, this indicates an inverse process. In this way, we can obtain the phase change of water in real time. In addition, from R2 in Eq. (6) we can extract some information about the aerosol droplet size when its size is greater than the laser wavelength. This theory, which was introduced by Whiteman and Melfi [9], has been tested by many scientists because the extinction and Liquid-Raman backscattering coefficients can be accepted as being dependent upon the liquid water area (r2 ) and volume (4r3 /3).
I. Song et al. / Optik 124 (2013) 1170–1174
1173
4. Discussion and conclusion
Fig. 4. Change in elastic, color ratio, and liquid water Raman/extinction coefficient ratio with altitude within the boundary layer and near the cloud measured on November 14, 2007 at different times; (a) the elastic, color ratio, and liquid water Raman/extinction coefficient ratio at 20:35–21:05 (KST), (b) the elastic, color ratio, and liquid water Raman/extinction coefficient ratio at 22:50–23:20 (KST).
ratio and liquid water Raman/extinction coefficient ratio can give information about change in particle size. The color ratio and liquid water Raman/extinction coefficient ratio show a similar pattern at the same altitude (1400–1800 m). The color ratio, however, shows different shapes (an increase) in the cloud, while the liquidwater Raman/extinction coefficient ratio decreased from the cloud during the measurement period, as shown in the figure. Therefore, it is necessary to distinguish between the aerosol and the cloud. The color ratio applies to the aerosol, while the liquid-water Raman/extinction coefficient ratio does applies to the cloud. In other words, the color ratio can give information about the size change of an aerosol, while the Raman/extinction coefficient ratio can give information about the size change of a cloud.
Fig. 5. Change in elastic, color ratio, and liquid water Raman/extinction coefficient ratio with altitude near cloud measured on August 20, 2007 (following a rainy day) at 21:10–22:25 (KST).
We have installed a multi-channel Raman/Mie lidar system that can characterize aerosols and cloud droplets. We have also obtained many kinds of vibrational and rotational Raman signals and Mie lidar signals in order to calculate the aerosol (cloud) backscattering and extinction coefficients. The spatial and temporal aerosol (cloud) extinction coefficient, color ratio, and the relative liquid-water content were measured using these multi-channel Raman lidar signals. Using these scattering coefficients, we calculated several aerosol parameters (R1 , R2 , reff , ˇ), and with these parameters we were able to determine the characteristics of the aerosol’s hygroscopic character, effective size, and water droplet effective size. As shown in Fig. 5, the liquid water Raman/extinction ratio decreased at the cloud because of differences in the aerosol and water droplet size distribution. That is to say, the scattering efficiency of the Mie and liquid-water Raman change dramatically, and also change differently when the aerosol effective size changes from a small value to large value: the scattering cross-section also changes in this size range. Thus, it is necessary to distinguish between the aerosol and the cloud. The color ratio method can be applied to the aerosol, while the liquid-water Raman/extinction coefficient ratio can be applied to the cloud droplet size measurements. We have also determined the aerosol’s hygroscopic characteristics. Another liquid water that is attached to the aerosol can also be measured when we measure the liquid water Raman lidar under normal atmospheric conditions and then normalized the obtained lidar signals using the water vapor Raman lidar. This suggested approach is not yet completely successful. However, if we measure these ratios continuously at the cloud and aerosol, and calibrate the effective size of the absolute cloud water droplet, we can measure the variations in the cloud water’s droplet size.
Acknowledgment This work was funded by the Korea Meteorological Administration Research (CATER 2006–3101) and the Korea Atomic Energy Research Institute.
References [1] Y. Ming, V. Ramaswamy, L.J. Donner, V.T.J. Phillips, Modeling the interactions between aerosols and liquid water clouds with a self-consistent cloud scheme in a general circulation model, J. Atmos. Sci. 64 (2006) 1189. [2] B. Tartarov, N. Sugimoto, I. Matsui, D.H. Shin, D. Muller, Multi-channel lidar spectrometer for atmospheric aerosol typing on the basis of chemical signatures in Raman spectra, in: 25th International Laser Radar Conference, 2010, pp. 47–50. [3] A. Shimizu, N. Sugimoto, I. Matsui, K. Arao, T. Murayama, N. Kagawa, K. Aoki, A. Uchiyama, A. Yamazaki, Continuous observations of Asian dust and other aerosols by polarization lidars in China and Japan during ACE-Asia, J. Geophys. Res. 109 (D19S17) (2004) 14. [4] Z. Liu, M. Vaughan, C. Kittaka, R. Kuehn, A. Omar, K. Powell, The CALIPSO lidar cloud and aerosol discrimination: version 2 – algorithm, performance and validation strategies, in: 24th International Lidar Radar Conference, 2008, pp. 1133–1136. [5] I. Uno, K. Yumimoto, A. Shimizu, Y. Hara, N. Sugimoto, Z. Wang, Z. Liu, D. Winker, 3D structure of Asian dust transport revealed by CALIPSO lidar and a 4D VAR dust model, Geophys. Res. Lett. 35 (2008) L06803. [6] W.K. Lee, H. Kil, Y.-S. Kwak, Q. Wu, S. Cho, J.U. Park, The winter anomaly in the middle-latitude F region during the solar minimum period observed by the constellation observing system for meteorology ionosphere and climatology, J. Geophys. Res. 116 (2010) 0148. [7] I.K. Song, Y.G. Kim, S.H. Baik, S.K. Park, H.G. Cha, S.C. Choi, C.M. Chung, Measurement of aerosol parameters with altitude by using two wavelength rotational Raman signals, J. Opt. Soc. Korea 14 (2010) 221–227. [8] D. Kim, S. Park, H.G. Cha, J. Zhou, W. Zhang, New multi-quantum number rotational Raman lidar for obtaining temperature and aerosol extinction and backscattering scattering coefficients, Appl. Phys. B 82 (2006) 1–4.
1174
I. Song et al. / Optik 124 (2013) 1170–1174
[9] D.N. Whiteman, S.H. Melfi, Cloud liquid water, mean droplet radius and number density measurements using a Raman lidar, J. Geophys. Res. 104 (1999) 31411. [10] I. Veselovskii, V. Griaznov, A. Kolgotin, D.N. Whiteman, Angle- and size-dependent characteristics of incoherent Raman and fluorescent scattering by microspheres. 2. Numerical simulation, Appl. Opt. 41 (2002) 5783. [11] A. Ansmann, M. Riebesell, U. Wandinger, C. Weitkamp, E. Voss, W. Lahmann, W. Michaelis, Combined Raman elastic-backscatter LIDAR for vertical profiling of moisture, aerosol extinction, backscatter, and LIDAR ratio, Appl. Phys. B 55 (1992) 18–28. [12] A. Ansmann, J. Bösenberg, A. Chaikovsky, A. Comerón, A. Stohl, O. Torres, G. Vaughan, W. Ulla, M. Wiegne, Long-range transport of Saharan dust to northern Europe: the 11–16 October 2001 outbreak observed with EARLINET, J. Geophys. Res. 108 (D24) (2003) 4783. [13] J.L. Tackett, L. Di Girolamo, Enhanced aerosol backscatter adjacent to tropical trade wind clouds revealed by satellite-based lidar, Geophys. Res. Lett. 36 (2009) L14804.
[14] B. de Foy, S.P. Burton, R.A. Ferrare, C.A. Hostetler, J.W. Hair, C. Wiedinmyer, L.T. Molina, Aerosol plume transport and transformation in high spectral resolution lidar measurements and WRF-Flexpart simulations during the MILAGRO field campaign, Atmos. Chem. Phys. 11 (2011) 3543–3563. [15] S.P. Burton, R.A. Ferrare, C.A. Hostetler, J.W. Hair, R.R. Rogers, M.D. Obland, C.F. Butler, A.L. Cook, D.B. Harper, K.D. Froyd, Aerosol classification using airborne high spectral resolution lidar measurements-methhodology and examples, Atmos. Meas. Tech. Discuss. 4 (2011) 5631–5688. [16] N. Sugimoto, C.H. Lee, Characteristics of dust aerosols inferred from lidar depolarization measurements at two wavelengths, Appl. Opt. 45 (2006) 7468–7474. [17] D. Kim, S.H. Baik, H.G. Cha, Y.G. kim, I.K. Song, Lidar measurement of a full Raman spectrum of water by using a multichannel detector, J. Korean Phys. Soc. 54 (2009) 38–43. [18] S.H. Park, D. Kim, H.D. Cheong, Y.G. Kim, Selecting characteristic Raman wavelengths to distinguish liquid water, water vapor, and ice water, J. Opt. Soc. Korea 14 (2010) 209. [19] D. Kim, H. Cha, Rotational Raman lidar for obtaining aerosol scattering coefficients, Opt. Lett. 30 (2005) 1728–1730.
Contents lists available at SciVerse ScienceDirect
Optik journal homepage: www.elsevier.de/ijleo
Aerosol and cloud characteristics analysis methods using multiple kinds of Raman lidar signals Imkang Song a , Il-Moon Chae a , Sung-Hoon Baik b , Sunngchul Choi b , Dukhyeon Kim c,∗ , Sunho Park c , Yonggi Kim d a
EngiOn Co., Ltd, 1-508, INNOPLEX, 552-2 Woncheon-dong, Yeongtong-gu, Suwon, 443-380, Republic of Korea Laboratory for Quantum Optics, Korea Atomic Energy Research Institute, Daejeon 305-353 Republic of Korea c Division of Cultural Studies, Hanbat National University, Daejeon 305-719, Republic of Korea d Department of Physics, Kongju National University, Kongju 314-701, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 15 October 2011 Accepted 4 March 2012
Keywords: Rotational Raman lidar Color Ratio Cloud Aerosol Liquid water
a b s t r a c t In this paper, we introduce three kinds of methods for the analysis of aerosol and cloud droplet characteristics: backscattering color ratio at two wavelengths (color ratio), aerosol liquid-water content, and cloud droplet size distribution using the liquid water and aerosol extinction coefficients. Based on theoretical perspective as well as our experimental results, we find that the liquid-water Raman scattering efficiency does not depend on four orders of particle size, but rather, depends on less than four orders, particularly in a smaller effective size distribution than Mie scattering efficiency. Therefore, we conclude that the color ratio method can be applied to aerosols, while the ratio between the liquid water Raman and extinction coefficients can be applied to cloud droplet size measurements. © 2012 Elsevier GmbH. All rights reserved.
1. Introduction Aerosol characteristics are an important parameter in long-term climate, and short-term weather changes because the aerosol scattering phase function and the physics of cloud formation depend upon aerosol size, shape, and aerosol chemical-optical characteristics. Also, a variety of natural and anthropogenic aerosol sources not only endow clouds with more reflectivity, but also lengthens the lifetime of clouds by suppressing precipitation [1]. However aerosol and liquid water parameters are known only on a local basis with low time resolutions and a limited range of characteristics, because most scientists have used traditional point-measuring equipments. Many researchers have tried to measure a variety of aerosol chemical and physical properties such as chemical components, physical characteristics, size distribution, and the refractive index by using a multi-wavelength Raman lidar and multiple-wavelength depolarization lidar. Tartarov et al. [2] provided a number of signatures of the chemical components of atmospheric aerosols using a multi-channel lidar spectrometer, but the spectrometer they used did not have a sufficient rough spectral resolution or spectral range. They also required too much time because of the low aerosol Raman scattering cross-section. Shimizu et al. [3] observed Asian dust and other aerosol characteristics using a polarization lidar. However,
∗ Corresponding author. E-mail address: [email protected] (D. Kim). 0030-4026/$ – see front matter © 2012 Elsevier GmbH. All rights reserved. doi:10.1016/j.ijleo.2012.03.012
the depolarization method is an indirect method, and therefore, has a complex dependence upon aerosol shape and aerosol alignment in the atmosphere. Thus, this method has a limited sphere of application. Furthermore, dust is an aerosol type that can be misclassified as a cloud when the dust layer is very dense [4]. Also Asian dust events are normally transported by cold fronts, and the transported dust is frequently adjacent to, embedded in, or mixed with clouds [5]. Satellite measuring systems can give global and 24 h information about aerosol extinction or backscattering coefficients, but they have number of limitations as regards spatial and time resolutions. Because these satellite data have limitations in obtaining certain important information about aerosol chemical or physical parameters and their altitude distributions, we need to find another method that can calibrate satellite data. From meteorological and health point of view, both the aerosol parameters and spatial distribution are important, so we must measure aerosol characteristics for a large spatial distribution as well. The concentration and profile of liquid water and water vapor can be measured by using a radiometer and a GNSS system with a low spatial resolution [6], but aerosol hygroscopic parameters cannot be measured via these methods. Cloud droplet size distribution depends upon the aerosol particle distribution and aerosol hygroscopic characteristics. In addition, aerosol atmospheric liquid water is composed of water droplets that are attached to the aerosol nucleus. To specify aerosol’s hygroscopic characteristics, we must also measure these characteristics.
I. Song et al. / Optik 124 (2013) 1170–1174
1171
Recently, other scientists have tried to combine more than two parameters in order to distinguish among aerosols by using approaches such as the lidar ratio, depolarization, and color ratio etc. [7,8]. If we can simultaneously measure a liquid water lidar signal, a nitrogen Raman lidar signal, and an aerosol Mie lidar signal, we can characterize an aerosol’s hygroscopic characteristics and its cloud droplet size. The main purpose of this study is to introduce such a method of measurement and to demonstrate it in terms of aerosol hygroscopic characteristics, cloud droplet size, and particle size, using liquid water Raman, rotational, and Mie lidar systems. 2. Method Fig. 1. Mie and liquid water Raman scattering efficient [10].
Although many scientists have tried to measure aerosols by using various lidar systems, their approaches have all treated water cloud droplets and aerosols simultaneously. However, cloud droplets have a constant refractive index and a specific size distribution that depend upon cloud height and other conditions. Here, we discuss aerosol hygroscopic characteristics, aerosol effective size distribution, and cloud droplet size distribution from a number of theoretical viewpoints.
smaller than the water droplet Raman efficiency, but as the droplet size increases, the Mie scattering efficiency increases more quickly than the Raman scattering. This means that when the liquid water lidar signal is normalized by an aerosol lidar signal, the normalized liquid water Raman values are dependent upon the aerosol droplet size.
2.1. Method for obtaining cloud droplet size distribution
2.2. Methods for obtaining aerosol size distributions
Mie scattering lidar signals contain information about the aerosol backscattering and extinction coefficients, while liquid water Raman lidar signals contain information about the total liquid water volume that is attached to the aerosol. Whiteman and Melfi [9] have suggested a simple method for measuring water droplet size. They were able to measure cloud droplet size using liquid water Raman lidar signals and extinction coefficients. As calculated by Veselovskii et al. [10], the Raman scattering efficiency of liquid water is constant (QLiquid Raman ≈ 4), and is approximately twice the extinction efficiency (Qext ≈ 2) when the size parameter x (2r/) is greater than 1. Hence, the total Raman signal of liquid water depends only upon the total volume of liquid water, as follows:
The backscatter ratio at different frequencies–the so called “color ratio”–depends upon the size distribution, the source of the aerosol (refractive index), and the shape of the aerosol particles, and is particularly useful for inferring the mean size. Ansmann et al. [12] suggest how the color ratio, P and S may be used to infer aerosol properties, and they offer validation of their approach as obtained from aircraft flights. Tackett et al. [13] also provide the first lidar observations of the backscatter-color ratio, and show that these quantities are directly related to aerosol properties. They show that the backscatter and color ratio are enhanced adjacent to the cloud edge, particularly near the cloud top and cloud base. Specifically, there is an increase of 31 ± 3% and 42 ± 2% in the layer-integrated median backscatter at wavelengths of 532 nm and 1064 nm, respectively, and the layer-averaged color ratio increases by 15 ± 5%. These backscatter calculations suggest that our observations mode adjacent to clouds are best explained by the aerosol size distribution with a reduced number concentration, increased median radius, and decreased width, as compared to observations made at a distance from clouds. B. de Foy et al. [14] also suggest four intensive parameters that depend only on the aerosol type and not on the concentration that can be calculated from the measurements: the lidar ratio, the aerosol depolarization ratios at 532 nm and 1064 nm, the backscatter coefficient, and the color ratio. Recently, Burton et al. [15] and Sugimoto et al. [16] suggested that although the depolarization color ratio has not been studied extensively, it can provide information about the relative size of non-spherical particles and may also provide a means of distinguishing between smoke and pollution.
cloud ˇLiquid
Raman
cloud n(r)QLiquid
=
≈
n(r)
Raman
(r)
4r 3 dr 3
16r 3 dr. 3
(1)
At the same time, the extinction coefficient depends upon the total area, as follows:
∞
n(r)Qext r 2 dr ≈
˛= 0
∞
n(r)2r 2 dr
(2)
0
Because we can easily measure a cloud’s optical and physical depth from the bottom of the cloud, we can easily determine a cloud’s extinction coefficient [7]. Consequently, from Eqs. (1) and (2), the effective size of the water droplet can be expressed as Eq. (3).
∞ cloud 3ˇLiquid /16 n(r)r 3 dr Raman reff = 0∞ = . 0
n(r)r 2 dr
˛/2
2.3. Method for obtaining aerosol hygroscopic characteristics (3)
cloud where ˇLiquid and ˛ are the water droplet cloud backscattering Raman coefficient and the extinction coefficient, respectively. These coefficients can be obtained using conventional lidar signal analysis methods [11]. Fig. 1 shows the liquid water Raman and Mie scattering efficiency of a water droplet [10]. As can be seen in the figure, when the aerosol droplet size is small, the Mie scattering efficiency is
Normally, atmospheric aerosol contains hydrogen-bonded water, a kind of water that has a different spectrum from aerosols that contain water vapor or bulk water. If we simultaneously measure the Raman signal of this bonded water and the aerosolscattering Mie signal, we can then extract useful information about the hygroscopic characteristics of the aerosol. For example, in a non-hygroscopic aerosol we find a strong Mie scattering signal even though we find a weak liquid water Raman lidar signal, and vice versa. When the Raman and Mie lidar receivers have the same lidar
1172
I. Song et al. / Optik 124 (2013) 1170–1174
Fig. 2. Schematic diagram of experimental system.
overlap function and measure the same direction, we can write the two lidar equations as follows:
3. Experimental setup and results
the rotational Raman signal, the nitrogen vibrational Raman signal for normalization, and a full spectrum (32 channels) of the vibrational Raman signal for water vapor/liquid water. This system can block out an elastic signal (355 nm) of more than 14 orders of magnitude and can resolve Raman wavelengths from 401 nm to 410 nm using 32-channel PMT sensors (Hamamatsu H7260A) [17,18]. From the 355 nm rotational Raman and 532 nm rotational Raman and elastic Mie signals, we can calculate the aerosol backscattering coefficient without any assumptions [19]. Fig. 3 shows the traditional elastic, rotational Raman, and the nitrogen rotational Raman lidar signals obtained at these two wavelengths. As can be seen in the figure, a strong elastic backscatter signal was measured in the presence of clouds at around 4.5 km, while the rotational Raman signal disappeared in this region. There was no leakage of the elastic signal into the Raman channel, even in densely cloudy conditions. Fig. 4(a) and (b) shows the changes in the of elastic, color ratio, and liquid water Raman/extinction ratio at different altitudes in the boundary layer near a cloud on November 14, 2007. The color ratio and liquid water Raman/extinction coefficient ratio can give information about the change in particle size. In the figures, it can be seen that even though the boundary layer is at the same altitude (1800 m), its range changes with time. As regards to the cloud conditions, the cloud’s bottom height was 2100 m at 20:35, but after the boundary layer collapsed, the bottom height became lower at 22:50 (1900 m) than it was at 20:35. In this case, the three parameters show very similar patterns in the mist (1400–1600 m), the boundary layer (1800 m), and the cloud (2200–2400 m). From these figures, we can deduce that the aerosol concentration and aerosol size distribution change. Fig. 5 shows the change in the elastic, color ratio, and liquid water Raman/extinction coefficient ratio near the cloud on August 20, 2007 (following a rainy day). This figure shows the changes in the elastic, color ratio, and liquid water Raman/extinction coefficient ratio near the cloud at different altitudes. The color
Fig. 2 shows a schematic diagram of the installed lidar system. A pulsed Nd:YAG laser was used as a transmitter and produced 6 ns pulses at a repetition rate of 30 Hz, which a maximum energy of 150 mJ/pulse at 355 nm, and 300 mJ at 532 nm. To decrease the degree of laser beam divergence, a beam expander was used in the transmitting system that could expand a laser beam 10 times. The two wavelength (355 nm, 532 nm) lidar signals obtained from two telescopes were the input to the fiber-based double-grating monochrometer. From the hardware perspective, the receiving system has three conceptual configurations. The first conceptual channel is water vapor, liquid water and ice water channels, and the second conceptual channel consists of rotational Raman channels for 355 nm and 532 nm. The third configuration is a Mie scattering channel. From the 355-nm wavelength laser, we can receive rotational Raman, elastic Mie, and water Raman signals. The receiving system consisted of four signals from as many detectors: the elastic signal,
Fig. 3. Traditional return signals for elastic, rotational Raman, and nitrogen rotational Raman LIDAR signals on October 12, 2007 (KST). The elastic signal was not affected in the rotational Raman channel.
PMie,Raman (z) = K(z)
ˇMie,Raman z2
e
−
z 0
˛(z ,0 )+˛(z ,Mie,Raman ) dz
.
(4)
Here, K, , ˇ, z, ˛, denote a constant, the geometric overlap function, the backscattering coefficient, distance (z), the extinction coefficient, and the wavelength for a given subscript, such as liquid water Raman (Liquid Raman, Vapor Raman), nitrogen molecule, and aerosol (Mie), and other Raman (R) wavelengths. When the liquid-water lidar signal is divided by the aerosol and water vapor Raman signals, the ratios are approximately written as follows: R1 (z) =
R2 (z) =
PLiquid Pvapor (z) PLiquid PMie (z)
,
=
(5) ˇRaman e ˇMie
z 0
˛(z ,Raman )−˛(z ,Mie ) dz
.
(6)
From R1 of Eq. (5) we can obtain some relative information about a liquid water cloud. An increase in the value (R1 ) time indicates an increase in the cloud droplet concentration or a decrease in the cloud water vapor. Therefore this value indicates the relative liquid water content of a given water vapor. Changes in this R1 indicate changes in the phase of water from vapor to liquid. If this value decreases, this indicates an inverse process. In this way, we can obtain the phase change of water in real time. In addition, from R2 in Eq. (6) we can extract some information about the aerosol droplet size when its size is greater than the laser wavelength. This theory, which was introduced by Whiteman and Melfi [9], has been tested by many scientists because the extinction and Liquid-Raman backscattering coefficients can be accepted as being dependent upon the liquid water area (r2 ) and volume (4r3 /3).
I. Song et al. / Optik 124 (2013) 1170–1174
1173
4. Discussion and conclusion
Fig. 4. Change in elastic, color ratio, and liquid water Raman/extinction coefficient ratio with altitude within the boundary layer and near the cloud measured on November 14, 2007 at different times; (a) the elastic, color ratio, and liquid water Raman/extinction coefficient ratio at 20:35–21:05 (KST), (b) the elastic, color ratio, and liquid water Raman/extinction coefficient ratio at 22:50–23:20 (KST).
ratio and liquid water Raman/extinction coefficient ratio can give information about change in particle size. The color ratio and liquid water Raman/extinction coefficient ratio show a similar pattern at the same altitude (1400–1800 m). The color ratio, however, shows different shapes (an increase) in the cloud, while the liquidwater Raman/extinction coefficient ratio decreased from the cloud during the measurement period, as shown in the figure. Therefore, it is necessary to distinguish between the aerosol and the cloud. The color ratio applies to the aerosol, while the liquid-water Raman/extinction coefficient ratio does applies to the cloud. In other words, the color ratio can give information about the size change of an aerosol, while the Raman/extinction coefficient ratio can give information about the size change of a cloud.
Fig. 5. Change in elastic, color ratio, and liquid water Raman/extinction coefficient ratio with altitude near cloud measured on August 20, 2007 (following a rainy day) at 21:10–22:25 (KST).
We have installed a multi-channel Raman/Mie lidar system that can characterize aerosols and cloud droplets. We have also obtained many kinds of vibrational and rotational Raman signals and Mie lidar signals in order to calculate the aerosol (cloud) backscattering and extinction coefficients. The spatial and temporal aerosol (cloud) extinction coefficient, color ratio, and the relative liquid-water content were measured using these multi-channel Raman lidar signals. Using these scattering coefficients, we calculated several aerosol parameters (R1 , R2 , reff , ˇ), and with these parameters we were able to determine the characteristics of the aerosol’s hygroscopic character, effective size, and water droplet effective size. As shown in Fig. 5, the liquid water Raman/extinction ratio decreased at the cloud because of differences in the aerosol and water droplet size distribution. That is to say, the scattering efficiency of the Mie and liquid-water Raman change dramatically, and also change differently when the aerosol effective size changes from a small value to large value: the scattering cross-section also changes in this size range. Thus, it is necessary to distinguish between the aerosol and the cloud. The color ratio method can be applied to the aerosol, while the liquid-water Raman/extinction coefficient ratio can be applied to the cloud droplet size measurements. We have also determined the aerosol’s hygroscopic characteristics. Another liquid water that is attached to the aerosol can also be measured when we measure the liquid water Raman lidar under normal atmospheric conditions and then normalized the obtained lidar signals using the water vapor Raman lidar. This suggested approach is not yet completely successful. However, if we measure these ratios continuously at the cloud and aerosol, and calibrate the effective size of the absolute cloud water droplet, we can measure the variations in the cloud water’s droplet size.
Acknowledgment This work was funded by the Korea Meteorological Administration Research (CATER 2006–3101) and the Korea Atomic Energy Research Institute.
References [1] Y. Ming, V. Ramaswamy, L.J. Donner, V.T.J. Phillips, Modeling the interactions between aerosols and liquid water clouds with a self-consistent cloud scheme in a general circulation model, J. Atmos. Sci. 64 (2006) 1189. [2] B. Tartarov, N. Sugimoto, I. Matsui, D.H. Shin, D. Muller, Multi-channel lidar spectrometer for atmospheric aerosol typing on the basis of chemical signatures in Raman spectra, in: 25th International Laser Radar Conference, 2010, pp. 47–50. [3] A. Shimizu, N. Sugimoto, I. Matsui, K. Arao, T. Murayama, N. Kagawa, K. Aoki, A. Uchiyama, A. Yamazaki, Continuous observations of Asian dust and other aerosols by polarization lidars in China and Japan during ACE-Asia, J. Geophys. Res. 109 (D19S17) (2004) 14. [4] Z. Liu, M. Vaughan, C. Kittaka, R. Kuehn, A. Omar, K. Powell, The CALIPSO lidar cloud and aerosol discrimination: version 2 – algorithm, performance and validation strategies, in: 24th International Lidar Radar Conference, 2008, pp. 1133–1136. [5] I. Uno, K. Yumimoto, A. Shimizu, Y. Hara, N. Sugimoto, Z. Wang, Z. Liu, D. Winker, 3D structure of Asian dust transport revealed by CALIPSO lidar and a 4D VAR dust model, Geophys. Res. Lett. 35 (2008) L06803. [6] W.K. Lee, H. Kil, Y.-S. Kwak, Q. Wu, S. Cho, J.U. Park, The winter anomaly in the middle-latitude F region during the solar minimum period observed by the constellation observing system for meteorology ionosphere and climatology, J. Geophys. Res. 116 (2010) 0148. [7] I.K. Song, Y.G. Kim, S.H. Baik, S.K. Park, H.G. Cha, S.C. Choi, C.M. Chung, Measurement of aerosol parameters with altitude by using two wavelength rotational Raman signals, J. Opt. Soc. Korea 14 (2010) 221–227. [8] D. Kim, S. Park, H.G. Cha, J. Zhou, W. Zhang, New multi-quantum number rotational Raman lidar for obtaining temperature and aerosol extinction and backscattering scattering coefficients, Appl. Phys. B 82 (2006) 1–4.
1174
I. Song et al. / Optik 124 (2013) 1170–1174
[9] D.N. Whiteman, S.H. Melfi, Cloud liquid water, mean droplet radius and number density measurements using a Raman lidar, J. Geophys. Res. 104 (1999) 31411. [10] I. Veselovskii, V. Griaznov, A. Kolgotin, D.N. Whiteman, Angle- and size-dependent characteristics of incoherent Raman and fluorescent scattering by microspheres. 2. Numerical simulation, Appl. Opt. 41 (2002) 5783. [11] A. Ansmann, M. Riebesell, U. Wandinger, C. Weitkamp, E. Voss, W. Lahmann, W. Michaelis, Combined Raman elastic-backscatter LIDAR for vertical profiling of moisture, aerosol extinction, backscatter, and LIDAR ratio, Appl. Phys. B 55 (1992) 18–28. [12] A. Ansmann, J. Bösenberg, A. Chaikovsky, A. Comerón, A. Stohl, O. Torres, G. Vaughan, W. Ulla, M. Wiegne, Long-range transport of Saharan dust to northern Europe: the 11–16 October 2001 outbreak observed with EARLINET, J. Geophys. Res. 108 (D24) (2003) 4783. [13] J.L. Tackett, L. Di Girolamo, Enhanced aerosol backscatter adjacent to tropical trade wind clouds revealed by satellite-based lidar, Geophys. Res. Lett. 36 (2009) L14804.
[14] B. de Foy, S.P. Burton, R.A. Ferrare, C.A. Hostetler, J.W. Hair, C. Wiedinmyer, L.T. Molina, Aerosol plume transport and transformation in high spectral resolution lidar measurements and WRF-Flexpart simulations during the MILAGRO field campaign, Atmos. Chem. Phys. 11 (2011) 3543–3563. [15] S.P. Burton, R.A. Ferrare, C.A. Hostetler, J.W. Hair, R.R. Rogers, M.D. Obland, C.F. Butler, A.L. Cook, D.B. Harper, K.D. Froyd, Aerosol classification using airborne high spectral resolution lidar measurements-methhodology and examples, Atmos. Meas. Tech. Discuss. 4 (2011) 5631–5688. [16] N. Sugimoto, C.H. Lee, Characteristics of dust aerosols inferred from lidar depolarization measurements at two wavelengths, Appl. Opt. 45 (2006) 7468–7474. [17] D. Kim, S.H. Baik, H.G. Cha, Y.G. kim, I.K. Song, Lidar measurement of a full Raman spectrum of water by using a multichannel detector, J. Korean Phys. Soc. 54 (2009) 38–43. [18] S.H. Park, D. Kim, H.D. Cheong, Y.G. Kim, Selecting characteristic Raman wavelengths to distinguish liquid water, water vapor, and ice water, J. Opt. Soc. Korea 14 (2010) 209. [19] D. Kim, H. Cha, Rotational Raman lidar for obtaining aerosol scattering coefficients, Opt. Lett. 30 (2005) 1728–1730.