2568
OPTICS LETTERS / Vol. 40, No. 11 / June 1, 2015
Near-infrared open-path measurement of CO2 concentration in the urban atmosphere Hayato Saito,1,* Naohiro Manago,1 Kenji Kuriyama,2 and Hiroaki Kuze1 1
Center for Environmental Remote Sensing (CEReS), Chiba University 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan 2 Faculty of Engineering, Shizuoka University, 3-5-1 Johoku, Naka-ku, Hamamatsu 432-8561, Japan *Corresponding author: haya_saito@chiba‑u.jp Received January 28, 2015; revised May 3, 2015; accepted May 11, 2015; posted May 12, 2015 (Doc. ID 233298); published May 25, 2015 Average concentration of carbon dioxide (CO2 ) has been measured over a path length of 5.1 km in the lower troposphere by the method of differential optical absorption spectroscopy (DOAS) using a near-infrared light source based on amplified spontaneous emission. The analysis of CO2 absorption intensity around 1575 nm observed during 10 days over the Chiba city area has revealed that the CO2 concentration varied in the range of around 360–450 ppmv, with presumable influence of air mass advection from nearby industrial facilities. In addition, a good correlation has been found in relative humidity values between the DOAS and meteorological station data. As a whole, the present result indicates the usefulness of such a DOAS approach for measuring the concentration of CO2 averaged over an optical path of a few kilometers in the lower troposphere. © 2015 Optical Society of America OCIS codes: (120.0280) Remote sensing and sensors; (280.4788) Optical sensing and sensors; (300.6320) Spectroscopy, high-resolution; (300.6340) Spectroscopy, infrared. http://dx.doi.org/10.1364/OL.40.002568
The influence of carbon dioxide (CO2 ) and other greenhouse gases has been discussed in the context of global climate change [1,2]. Since the discovery of increasing trend in CO2 concentration at the Mauna Loa Observatory in Hawaii [3], sampling (flask) measurements have been undertaken in a number of points worldwide [4]. Aircraft measurement [5], as well as satellite measurements [6,7], has been undertaken for monitoring greenhouse gases on a routine basis. In addition, the feasibility of standoff measurement of CO2 concentration in the troposphere has been demonstrated by use of active remote sensing instruments, such as a differential absorption lidar (DIAL) [8] and a differential optical absorption spectroscopy (DOAS) system with a nanosecond white light continuum [9]. Because of the diversity and spatial distributions of CO2 emission sources, the results of sampling measurements do not always represent the regionally averaged CO2 concentration. Currently, fossil fuel consumption data are combined with atmospheric transport model simulations to estimate the spatial distribution of CO2 [10]. In this context, it is highly desirable to realize the concentration measurement in a wide urban area, where various anthropogenic sources are located. In addition, such remote sensing methodology covering wide areas would be suitable for the measurement of biospheric and oceanic fluxes of carbon dioxide. Visible or near ultraviolet spectral regions have usually been exploited for conventional DOAS measurements of pollutant gases [11–15]. In contrast, the present Letter describes the application of near-infrared (NIR) spectroscopy for the DOAS-type open path measurement of CO2 concentration in an urban environment. In a similar NIR measurement of CO2 around 2000 nm using a nanosecond white light continuum [9], the monitoring was made over one night with a spectral resolution of 8 nm. A much better resolution (0.046 nm) in the present work has enabled the clear separation of each rovibrational spectral line, and concentration changes could be measured both daytime and nighttime during a 10-day period. To the best 0146-9592/15/112568-04$15.00/0
of our knowledge, this is the first case that such an NIR DOAS approach has been applied for such long-term observation over the urban canopy. The light path was over the central part of Chiba city, a city inside the large Tokyo metropolitan area along the eastern coast of Tokyo Bay. The path length was 5.10 km, composed of a round trip (2.55 km) between Chiba University (35.63°N, 140.10°E) and the Chiba City Science Museum (35.61°N, 140.12°E). The average height of the light path was approximately 30 m above the ground level (50 m above sea level). There are industrial and forest areas on the western and eastern side of the light path, respectively, whereas the ground coverage beneath the light path is mostly buildings and streets with limited vegetation coverage. In addition to the heavy traffic in the city center, presumable CO2 sources are found in an industrial area along the coastal zone, including steel mills and two liquefied natural gas power plants. Figure 1 shows the experimental setup used for the present open-path DOAS measurement. A guide laser (532 nm) placed in the close vicinity of the transmitting telescope is used to attain the optical alignment of the telescopes. A C L band amplified spontaneous emission light source (Fiber Labs Inc., ASE-FL-7015) is employed for obtaining broadband output spectrum between 1530 and 1610 nm with a total output power
Fig. 1. Experimental setup for the DOAS measurement. © 2015 Optical Society of America
June 1, 2015 / Vol. 40, No. 11 / OPTICS LETTERS
of ∼200 mW. The fiber output of this light source is connected to a Newtonian telescope (Vixen, R130Sf, 130 mm diameter). The NIR light beam is nearly collimated with an angular spread of ∼3 mrad, which results in a positional tolerance of ∼10 m at 2.55 km. Once the alignment is optimized, the system can be operated mostly without any readjustment. At dawn, however, frequent adjustment is often required, probably because of the thermal distortion of the buildings due to sunlight irradiation. After the nearly horizontal transmission of 2.55 km, the beam is reflected with a retro-reflector setup having an effective area of 136.8 cm2 . The angular accuracy of each reflector is on the order of 0.05 mrad. The light intensity received with a 200 mm diameter telescope (Vixen, R200S) is coupled via an optical fiber to a spectrometer (HORIBA, iHR-550) equipped with a liquidnitrogen cooled InGaAs array sensor. The spectral resolution of 0.046 nm full width at half maximum is attained when the slit width is set at 25 μm. Owing to the narrow acceptance angle of the receiver telescope, the signal-tonoise ratio (for shot noise) of a single CO2 absorption line was on the order of 50–150, depending on the extinction of NIR beam intensity because of tropospheric aerosol. The data analysis procedure is similar to the case of standard DOAS analysis [11–13]. Only the outline of the derivation process is described here. The signal detected with the array detector can be expressed as I obs λ kλI 0 λT m λT a λT gas λ;
(1)
where I obs λ is the observed spectrum (λ is the wavelength), kλ is the instrumental efficiency, I 0 λ is the initially transmitted spectrum, and T i λ (i m; a, and gas) stands for the transmittance due to air molecules, aerosol, and absorption of CO2 and water vapor (H2 O and HDO). Each transmittance factor, T i λ can be expressed in terms of the corresponding optical thickness (OT). The OT for molecular or aerosol extinction, τi λ (i m or a), is a slowly varying function of wavelength. For absorbing molecules, the Lambert–Beer law yields τgas λ Σj N j Lσ j λ, where N j is the number density of each gas species, L is the optical path length, and σ j λ is the absorption cross sections (j CO2 , H2 O and HDO), which are rapidly varying functions with wavelength. Thus, by taking the logarithm of Eq. (1) and applying the high-pass filtering, one obtains
2569
information provided in the HITRAN database [17], under assumed values of atmospheric pressure and temperature. Generally, the concentration of HDO is indicated by the variation of deuterium abundance ratio in per mill (δD) [18]: δD 1000 ×
R RSMOW
−1 ;
(3)
where R HDO∕H2 O is the ratio of the volume mixing ratio of HDO and H2 O, and RSMOW is the R value of Vienna standard mean ocean water (SMOW) [19]. The data presented here are taken from the open-path observation conducted in Chiba city from September 9 to 18, 2014. During this 10-day period, the weather was mostly sunny or cloudy, without noticeable rainfall except on September 11. The lowest and highest temperatures recorded at the Chiba Observatory of Japan Meteorological Agency (JMA) (35.60°N, 140.10°E, 4 m above sea level) were 18.8°C and 28.3°C, respectively. Although the DOAS equipment was operated continuously, the occurrence of morning haze sometimes interrupted the measurement because of the large extinction along the optical path. The acquisition duration of a single data point was 60–300 s, depending on the atmospheric conditions including the effect of atmospheric turbulence [20]. Below, the analysis is based on the spectrum averaged over 1 h. The values for atmospheric pressure and temperature were obtained from the JMA station data. Figure 2 shows an example of the DOAS spectral fitting for the data observed around 10:00 Japan standard time (JST) on September 13, 2014. It is noted that, since the DOAS analysis is based on the ratio of the observed and reference spectra I 0 λ∕I obs λ in Eq. 2], the wavelength sensitivity of the optical instrumentation can almost be eliminated. The reference spectrum, I 0 λ, was obtained by putting the receiving telescope in front of the transmitting telescope in the laboratory. Figure 2 indicates that the separation of three gas species has
X I λ ln kλ Δτλ ≡ F HP ln 0 N j LΔσ j λ: (2) I obs λ j Here Δτλ is the quantity usually referred to as differential OT. This result indicates that the removal of the baseline that varies with the wavelength only slowly can lead to the determination of N j , the number concentration of j-th gas species averaged over the observation path. In the analysis of the DOAS spectra, the discrete Fourier transform is applied for implementing the high-pass filtering. A sixth-order polynomial is employed for the baseline fitting, and the Levenberg–Marquardt algorithm for nonlinear least squares fitting [16]. In the spectral range between 1570 and 1590 nm, the functional form of σ j λ for each of these gas species can be calculated using the
Fig. 2. Example of DOAS analysis for the retrieval of gas concentrations: (a) CO2 ; (b) H2 O; (c) HDO; and (d) final residual of the nonlinear fitting.
2570
OPTICS LETTERS / Vol. 40, No. 11 / June 1, 2015
been achieved successfully in the present nonlinear analysis of the DOAS spectrum. Figures 3(a)–3(d) summarize the result of the DOAS analysis. Figure 3(a) shows the temperature variation and weather records for the 10-day period. It is noted that a typhoon passed off the eastern coast of Japan around 8 September 2014, resulting in the strengthening of the autumnal rain front. Figure 3(b) shows the comparison of humidity between the JMA observation and DOAS retrieval. The temporal change of DOAS data agrees well with that of the station data, which were recorded at an observatory located near the Tokyo Bay. Because of the horizontal distance and the height difference between the observatory and the DOAS light path, one cannot expect the complete agreement between these two independent data. Still, the result in Fig. 3(b) suggests the homogeneity of humidity in the Chiba city region during the observation period in September 2014. Figure 3(c) shows that the CO2 concentration changes in a range of around 360–450 ppmv. Although the CO2 concentration exhibits a slight indication of diurnal variation as found in relative humidity, shown in Fig. 3(b), more remarkable changes tend to appear in relation to the wind direction: namely, higher concentrations of CO2 are seen when the wind direction is southwestern to western, while relatively lower concentrations set in when eastern to southeastern winds prevail. This infers the possible relation between the CO2 concentration and air mass advection, which is consistent with the spatial configuration of the present DOAS path against the industrial complex and forest areas. Nevertheless, it is likely that the direct correlation between the wind direction and CO2 concentration is often obscured because of the presence of heavy traffic in the region, in addition to more local and frequent changes in wind speed and direction. For the most case in the present analysis, the standard
deviation of CO2 concentration is ∼0.5%, though we have the additional uncertainty in the temperature estimated from the JMA data. This uncertainty leads to ∼1% error in the retrieval of CO2 concentration. Figure 3(d) shows the temporal variation of δD, indicating the change in HDO concentration. The deuterium concentration of natural water changes in the range of 105 79 ppmv [19]. In spite of such a small value of the isotope ratio, the optical thickness of absorption lines of HDO is not negligibly small as compared with H2 O absorption: this is because of the larger absorption cross section of HDO than H2 O. Thus, the contribution of HDO has to be taken into account in the nonlinear fitting procedure, though the accuracy of concentration retrieval of HDO is less than the corresponding values of CO2 or H2 O. During the observation period, the value of δD varied between around −300‰ and 100‰, with the average value of −140‰. Further study would be needed to verify whether such values are reasonable in the lower troposphere when the actual isotopologue ratio is considered for water in the Tokyo Bay area. In summary, we have demonstrated the DOAS measurement in the NIR spectral region. The concentrations of CO2 and water vapor have been measured in the lower troposphere over an optical path length of 5.10 km. Nearly continuous data have been obtained and the optical thickness due to molecular absorption was successfully separated from the background extinction because of molecular and aerosol scattering. During the 10-day observation period reported here, we observed the variation of CO2 concentration in a range of 360–450 ppmv. This concentration change was partly ascribable to the change in wind direction, suggesting the contributions from regional industrial sources. It has also been found that the relative humidity derived from the water vapor concentration concurrently determined from the DOAS analysis agreed well with the value provided from a nearby JMA station. As compared with the conventional flask measurement, the advantage of the DOAS approach is the capability of providing data averaged over a path length of a few kilometers. Such regional values will be useful for evaluating the influence of various sources and sinks around the region, providing the observational basis for the source/sink analysis of CO2 . In addition, the NIR DOAS approach will be useful for monitoring other trace gas species such as carbon monoxide or methane, depending on their concentrations in atmosphere as compared with the optical path length available for the measurement. The authors would like to thank Prof. R. Oda of the Chiba Institute of Technology for her contribution during the initial stage of the present experiment. They are also grateful to Dr. K. Ohtaka, the director of Chiba City Science Museum, for his encouragement and permission to use the rooftop space for our observation.
Fig. 3. Results of DOAS analysis with meteorological data observed during September 8–18, 2014: (a) temperature with weather records, (b) humidity from the JMA station and DOAS retrieval, (c) CO2 concentration, (d) δD, indicating the change in HDO concentration, and (e) wind direction and speed observed at the JMA station.
References 1. IPCC AR5 (2014), http://www.ipcc.ch/index.htm. 2. D. S. Jenkinson, D. E. Adams, and A. Wild, Nature 351, 304 (1991). 3. R. B. Bacastow, C. D. Keeling, and T. P. Whorf, J. Geophys. Res. 90, 10529 (1985).
June 1, 2015 / Vol. 40, No. 11 / OPTICS LETTERS 4. K. A. Masarie, R. L. Langenfelds, C. E. Allison, T. J. Conway, E. J. Dlugokencky, R. J. Francey, P. C. Novelli, L. P. Steele, P. P. Tans, B. Vaughn, and J. W. C. White, J. Geophys. Res. 106, 445 (2001). 5. H. Matsueda, T. Machida, Y. Sawa, Y. Nakagawa, K. Hirotani, H. Ikeda, N. Kondo, and K. Goto, Pap. Meteorol. Geophys. 59, 1 (2008). 6. T. Yokota, Y. Yoshida, N. Eguchi, Y. Ota, T. Tanaka, H. Watanabe, and S. Maksyutov, SOLA 5, 160 (2009). 7. M. Inoue, I. Morino, O. Uchino, Y. Miyamoto, Y. Yoshida, T. Yokota, T. Machida, Y. Sawa, H. Matsueda, C. Sweeney, P. P. Tans, A. E. Andrews, S. C. Biraud, T. Tanaka, S. Kawakami, and P. K. Patra, Atmos. Chem. Phys. 13, 9771 (2013). 8. D. Sakaizawa, C. Nagasawa, T. Nogai, M. Abo, Y. Shibata, M. Nakazato, and T. Sakai, Appl. Opt. 48, 748 (2009). 9. T. Somekawa, N. Manago, H. Kuze, and M. Fujita, Opt. Lett. 36, 4782 (2011). 10. M. Satho, T. Matsuno, H. Tomita, H. Miura, T. Nasuno, and S. Iga, J. Comput. Phys. 227, 3486 (2008). 11. U. Platt and J. Stutz, Differential Optical Absorption Spectroscopy Principles and Applications (SpringerVerlag, 2008). 12. H. Saito, Y. Goto, Y. Mabuchi, I. Alimuddin, G. Bagtasa, N. Manago, H. Irie, I. Harada, T. Ishibashi, K. Yashiro, S. Kameyama, and H. Kuze, Open J. Air Pollut. 3, 20 (2014).
2571
13. Y. Kambe, Y. Yoshii, K. Takahashi, and K. Tonokura, J. Environ. Monit. 14, 944 (2012). 14. M. Buchwitz, R. de Beek, J. P. Burrows, H. Bovensmann, T. Warneke, J. Notholt, J. F. Meirink, A. P. H. Goede, P. Bergamaschi, S. Korner, M. Heimann, and A. Schulz, Atmos. Chem. Phys. 5, 941 (2005). 15. C. Kern, S. Trick, B. Rippel, and U. Platt, Appl. Opt. 45, 2077 (2006). 16. D. W. Marquardt, J. Soc. Ind. Appl. Math. 11, 431 (1963). 17. L. S. Rothman, I. E. Gordon, Y. Babikov, A. Barbe, D. Chris Benner, P. F. Bernath, M. Birk, L. Bizzocchi, V. Boudon, L. R. Brown, A. Campargue, K. Chance, L. Coudert, V. M. Devi, B. J. Drouin, A. Fayt, J. M. Flaud, R. R. Gamache, J. Harrison, J. M. Hartmann, C. Hill, J. T. Hodges, D. Jacquemart, A. Jolly, J. Lamouroux, R. J. LeRoy, G. Li, D. Long, C. J. Mackie, S. T. Massie, S. Mikhailenko, H. S. P. Müller, O. V. Naumenko, A. V. Nikitin, J. Orphal, V. Perevalov, A. Perrin, E. R. Polovtseva, C. Richard, M. A. H. Smith, E. Starikova, K. Sung, S. Tashkun, J. Tennyson, G. C. Toon, V. G. Tyuterev, J. Vander Auwera, and G. Wagner, J. Quant. Spectrosc. Radiat. Transfer 130, 4 (2013). 18. W. Dansgaard, Tellus 16, 436 (1964). 19. M. Berglund and M. E. Wieser, Pure Appl. Chem. 83, 397 (2011). 20. A. Mahalov and M. Moustaoui, Laser Photon. Rev. 4, 144 (2010).
OPTICS LETTERS / Vol. 40, No. 11 / June 1, 2015
Near-infrared open-path measurement of CO2 concentration in the urban atmosphere Hayato Saito,1,* Naohiro Manago,1 Kenji Kuriyama,2 and Hiroaki Kuze1 1
Center for Environmental Remote Sensing (CEReS), Chiba University 1-33 Yayoi-cho, Inage-ku, Chiba 263-8522, Japan 2 Faculty of Engineering, Shizuoka University, 3-5-1 Johoku, Naka-ku, Hamamatsu 432-8561, Japan *Corresponding author: haya_saito@chiba‑u.jp Received January 28, 2015; revised May 3, 2015; accepted May 11, 2015; posted May 12, 2015 (Doc. ID 233298); published May 25, 2015 Average concentration of carbon dioxide (CO2 ) has been measured over a path length of 5.1 km in the lower troposphere by the method of differential optical absorption spectroscopy (DOAS) using a near-infrared light source based on amplified spontaneous emission. The analysis of CO2 absorption intensity around 1575 nm observed during 10 days over the Chiba city area has revealed that the CO2 concentration varied in the range of around 360–450 ppmv, with presumable influence of air mass advection from nearby industrial facilities. In addition, a good correlation has been found in relative humidity values between the DOAS and meteorological station data. As a whole, the present result indicates the usefulness of such a DOAS approach for measuring the concentration of CO2 averaged over an optical path of a few kilometers in the lower troposphere. © 2015 Optical Society of America OCIS codes: (120.0280) Remote sensing and sensors; (280.4788) Optical sensing and sensors; (300.6320) Spectroscopy, high-resolution; (300.6340) Spectroscopy, infrared. http://dx.doi.org/10.1364/OL.40.002568
The influence of carbon dioxide (CO2 ) and other greenhouse gases has been discussed in the context of global climate change [1,2]. Since the discovery of increasing trend in CO2 concentration at the Mauna Loa Observatory in Hawaii [3], sampling (flask) measurements have been undertaken in a number of points worldwide [4]. Aircraft measurement [5], as well as satellite measurements [6,7], has been undertaken for monitoring greenhouse gases on a routine basis. In addition, the feasibility of standoff measurement of CO2 concentration in the troposphere has been demonstrated by use of active remote sensing instruments, such as a differential absorption lidar (DIAL) [8] and a differential optical absorption spectroscopy (DOAS) system with a nanosecond white light continuum [9]. Because of the diversity and spatial distributions of CO2 emission sources, the results of sampling measurements do not always represent the regionally averaged CO2 concentration. Currently, fossil fuel consumption data are combined with atmospheric transport model simulations to estimate the spatial distribution of CO2 [10]. In this context, it is highly desirable to realize the concentration measurement in a wide urban area, where various anthropogenic sources are located. In addition, such remote sensing methodology covering wide areas would be suitable for the measurement of biospheric and oceanic fluxes of carbon dioxide. Visible or near ultraviolet spectral regions have usually been exploited for conventional DOAS measurements of pollutant gases [11–15]. In contrast, the present Letter describes the application of near-infrared (NIR) spectroscopy for the DOAS-type open path measurement of CO2 concentration in an urban environment. In a similar NIR measurement of CO2 around 2000 nm using a nanosecond white light continuum [9], the monitoring was made over one night with a spectral resolution of 8 nm. A much better resolution (0.046 nm) in the present work has enabled the clear separation of each rovibrational spectral line, and concentration changes could be measured both daytime and nighttime during a 10-day period. To the best 0146-9592/15/112568-04$15.00/0
of our knowledge, this is the first case that such an NIR DOAS approach has been applied for such long-term observation over the urban canopy. The light path was over the central part of Chiba city, a city inside the large Tokyo metropolitan area along the eastern coast of Tokyo Bay. The path length was 5.10 km, composed of a round trip (2.55 km) between Chiba University (35.63°N, 140.10°E) and the Chiba City Science Museum (35.61°N, 140.12°E). The average height of the light path was approximately 30 m above the ground level (50 m above sea level). There are industrial and forest areas on the western and eastern side of the light path, respectively, whereas the ground coverage beneath the light path is mostly buildings and streets with limited vegetation coverage. In addition to the heavy traffic in the city center, presumable CO2 sources are found in an industrial area along the coastal zone, including steel mills and two liquefied natural gas power plants. Figure 1 shows the experimental setup used for the present open-path DOAS measurement. A guide laser (532 nm) placed in the close vicinity of the transmitting telescope is used to attain the optical alignment of the telescopes. A C L band amplified spontaneous emission light source (Fiber Labs Inc., ASE-FL-7015) is employed for obtaining broadband output spectrum between 1530 and 1610 nm with a total output power
Fig. 1. Experimental setup for the DOAS measurement. © 2015 Optical Society of America
June 1, 2015 / Vol. 40, No. 11 / OPTICS LETTERS
of ∼200 mW. The fiber output of this light source is connected to a Newtonian telescope (Vixen, R130Sf, 130 mm diameter). The NIR light beam is nearly collimated with an angular spread of ∼3 mrad, which results in a positional tolerance of ∼10 m at 2.55 km. Once the alignment is optimized, the system can be operated mostly without any readjustment. At dawn, however, frequent adjustment is often required, probably because of the thermal distortion of the buildings due to sunlight irradiation. After the nearly horizontal transmission of 2.55 km, the beam is reflected with a retro-reflector setup having an effective area of 136.8 cm2 . The angular accuracy of each reflector is on the order of 0.05 mrad. The light intensity received with a 200 mm diameter telescope (Vixen, R200S) is coupled via an optical fiber to a spectrometer (HORIBA, iHR-550) equipped with a liquidnitrogen cooled InGaAs array sensor. The spectral resolution of 0.046 nm full width at half maximum is attained when the slit width is set at 25 μm. Owing to the narrow acceptance angle of the receiver telescope, the signal-tonoise ratio (for shot noise) of a single CO2 absorption line was on the order of 50–150, depending on the extinction of NIR beam intensity because of tropospheric aerosol. The data analysis procedure is similar to the case of standard DOAS analysis [11–13]. Only the outline of the derivation process is described here. The signal detected with the array detector can be expressed as I obs λ kλI 0 λT m λT a λT gas λ;
(1)
where I obs λ is the observed spectrum (λ is the wavelength), kλ is the instrumental efficiency, I 0 λ is the initially transmitted spectrum, and T i λ (i m; a, and gas) stands for the transmittance due to air molecules, aerosol, and absorption of CO2 and water vapor (H2 O and HDO). Each transmittance factor, T i λ can be expressed in terms of the corresponding optical thickness (OT). The OT for molecular or aerosol extinction, τi λ (i m or a), is a slowly varying function of wavelength. For absorbing molecules, the Lambert–Beer law yields τgas λ Σj N j Lσ j λ, where N j is the number density of each gas species, L is the optical path length, and σ j λ is the absorption cross sections (j CO2 , H2 O and HDO), which are rapidly varying functions with wavelength. Thus, by taking the logarithm of Eq. (1) and applying the high-pass filtering, one obtains
2569
information provided in the HITRAN database [17], under assumed values of atmospheric pressure and temperature. Generally, the concentration of HDO is indicated by the variation of deuterium abundance ratio in per mill (δD) [18]: δD 1000 ×
R RSMOW
−1 ;
(3)
where R HDO∕H2 O is the ratio of the volume mixing ratio of HDO and H2 O, and RSMOW is the R value of Vienna standard mean ocean water (SMOW) [19]. The data presented here are taken from the open-path observation conducted in Chiba city from September 9 to 18, 2014. During this 10-day period, the weather was mostly sunny or cloudy, without noticeable rainfall except on September 11. The lowest and highest temperatures recorded at the Chiba Observatory of Japan Meteorological Agency (JMA) (35.60°N, 140.10°E, 4 m above sea level) were 18.8°C and 28.3°C, respectively. Although the DOAS equipment was operated continuously, the occurrence of morning haze sometimes interrupted the measurement because of the large extinction along the optical path. The acquisition duration of a single data point was 60–300 s, depending on the atmospheric conditions including the effect of atmospheric turbulence [20]. Below, the analysis is based on the spectrum averaged over 1 h. The values for atmospheric pressure and temperature were obtained from the JMA station data. Figure 2 shows an example of the DOAS spectral fitting for the data observed around 10:00 Japan standard time (JST) on September 13, 2014. It is noted that, since the DOAS analysis is based on the ratio of the observed and reference spectra I 0 λ∕I obs λ in Eq. 2], the wavelength sensitivity of the optical instrumentation can almost be eliminated. The reference spectrum, I 0 λ, was obtained by putting the receiving telescope in front of the transmitting telescope in the laboratory. Figure 2 indicates that the separation of three gas species has
X I λ ln kλ Δτλ ≡ F HP ln 0 N j LΔσ j λ: (2) I obs λ j Here Δτλ is the quantity usually referred to as differential OT. This result indicates that the removal of the baseline that varies with the wavelength only slowly can lead to the determination of N j , the number concentration of j-th gas species averaged over the observation path. In the analysis of the DOAS spectra, the discrete Fourier transform is applied for implementing the high-pass filtering. A sixth-order polynomial is employed for the baseline fitting, and the Levenberg–Marquardt algorithm for nonlinear least squares fitting [16]. In the spectral range between 1570 and 1590 nm, the functional form of σ j λ for each of these gas species can be calculated using the
Fig. 2. Example of DOAS analysis for the retrieval of gas concentrations: (a) CO2 ; (b) H2 O; (c) HDO; and (d) final residual of the nonlinear fitting.
2570
OPTICS LETTERS / Vol. 40, No. 11 / June 1, 2015
been achieved successfully in the present nonlinear analysis of the DOAS spectrum. Figures 3(a)–3(d) summarize the result of the DOAS analysis. Figure 3(a) shows the temperature variation and weather records for the 10-day period. It is noted that a typhoon passed off the eastern coast of Japan around 8 September 2014, resulting in the strengthening of the autumnal rain front. Figure 3(b) shows the comparison of humidity between the JMA observation and DOAS retrieval. The temporal change of DOAS data agrees well with that of the station data, which were recorded at an observatory located near the Tokyo Bay. Because of the horizontal distance and the height difference between the observatory and the DOAS light path, one cannot expect the complete agreement between these two independent data. Still, the result in Fig. 3(b) suggests the homogeneity of humidity in the Chiba city region during the observation period in September 2014. Figure 3(c) shows that the CO2 concentration changes in a range of around 360–450 ppmv. Although the CO2 concentration exhibits a slight indication of diurnal variation as found in relative humidity, shown in Fig. 3(b), more remarkable changes tend to appear in relation to the wind direction: namely, higher concentrations of CO2 are seen when the wind direction is southwestern to western, while relatively lower concentrations set in when eastern to southeastern winds prevail. This infers the possible relation between the CO2 concentration and air mass advection, which is consistent with the spatial configuration of the present DOAS path against the industrial complex and forest areas. Nevertheless, it is likely that the direct correlation between the wind direction and CO2 concentration is often obscured because of the presence of heavy traffic in the region, in addition to more local and frequent changes in wind speed and direction. For the most case in the present analysis, the standard
deviation of CO2 concentration is ∼0.5%, though we have the additional uncertainty in the temperature estimated from the JMA data. This uncertainty leads to ∼1% error in the retrieval of CO2 concentration. Figure 3(d) shows the temporal variation of δD, indicating the change in HDO concentration. The deuterium concentration of natural water changes in the range of 105 79 ppmv [19]. In spite of such a small value of the isotope ratio, the optical thickness of absorption lines of HDO is not negligibly small as compared with H2 O absorption: this is because of the larger absorption cross section of HDO than H2 O. Thus, the contribution of HDO has to be taken into account in the nonlinear fitting procedure, though the accuracy of concentration retrieval of HDO is less than the corresponding values of CO2 or H2 O. During the observation period, the value of δD varied between around −300‰ and 100‰, with the average value of −140‰. Further study would be needed to verify whether such values are reasonable in the lower troposphere when the actual isotopologue ratio is considered for water in the Tokyo Bay area. In summary, we have demonstrated the DOAS measurement in the NIR spectral region. The concentrations of CO2 and water vapor have been measured in the lower troposphere over an optical path length of 5.10 km. Nearly continuous data have been obtained and the optical thickness due to molecular absorption was successfully separated from the background extinction because of molecular and aerosol scattering. During the 10-day observation period reported here, we observed the variation of CO2 concentration in a range of 360–450 ppmv. This concentration change was partly ascribable to the change in wind direction, suggesting the contributions from regional industrial sources. It has also been found that the relative humidity derived from the water vapor concentration concurrently determined from the DOAS analysis agreed well with the value provided from a nearby JMA station. As compared with the conventional flask measurement, the advantage of the DOAS approach is the capability of providing data averaged over a path length of a few kilometers. Such regional values will be useful for evaluating the influence of various sources and sinks around the region, providing the observational basis for the source/sink analysis of CO2 . In addition, the NIR DOAS approach will be useful for monitoring other trace gas species such as carbon monoxide or methane, depending on their concentrations in atmosphere as compared with the optical path length available for the measurement. The authors would like to thank Prof. R. Oda of the Chiba Institute of Technology for her contribution during the initial stage of the present experiment. They are also grateful to Dr. K. Ohtaka, the director of Chiba City Science Museum, for his encouragement and permission to use the rooftop space for our observation.
Fig. 3. Results of DOAS analysis with meteorological data observed during September 8–18, 2014: (a) temperature with weather records, (b) humidity from the JMA station and DOAS retrieval, (c) CO2 concentration, (d) δD, indicating the change in HDO concentration, and (e) wind direction and speed observed at the JMA station.
References 1. IPCC AR5 (2014), http://www.ipcc.ch/index.htm. 2. D. S. Jenkinson, D. E. Adams, and A. Wild, Nature 351, 304 (1991). 3. R. B. Bacastow, C. D. Keeling, and T. P. Whorf, J. Geophys. Res. 90, 10529 (1985).
June 1, 2015 / Vol. 40, No. 11 / OPTICS LETTERS 4. K. A. Masarie, R. L. Langenfelds, C. E. Allison, T. J. Conway, E. J. Dlugokencky, R. J. Francey, P. C. Novelli, L. P. Steele, P. P. Tans, B. Vaughn, and J. W. C. White, J. Geophys. Res. 106, 445 (2001). 5. H. Matsueda, T. Machida, Y. Sawa, Y. Nakagawa, K. Hirotani, H. Ikeda, N. Kondo, and K. Goto, Pap. Meteorol. Geophys. 59, 1 (2008). 6. T. Yokota, Y. Yoshida, N. Eguchi, Y. Ota, T. Tanaka, H. Watanabe, and S. Maksyutov, SOLA 5, 160 (2009). 7. M. Inoue, I. Morino, O. Uchino, Y. Miyamoto, Y. Yoshida, T. Yokota, T. Machida, Y. Sawa, H. Matsueda, C. Sweeney, P. P. Tans, A. E. Andrews, S. C. Biraud, T. Tanaka, S. Kawakami, and P. K. Patra, Atmos. Chem. Phys. 13, 9771 (2013). 8. D. Sakaizawa, C. Nagasawa, T. Nogai, M. Abo, Y. Shibata, M. Nakazato, and T. Sakai, Appl. Opt. 48, 748 (2009). 9. T. Somekawa, N. Manago, H. Kuze, and M. Fujita, Opt. Lett. 36, 4782 (2011). 10. M. Satho, T. Matsuno, H. Tomita, H. Miura, T. Nasuno, and S. Iga, J. Comput. Phys. 227, 3486 (2008). 11. U. Platt and J. Stutz, Differential Optical Absorption Spectroscopy Principles and Applications (SpringerVerlag, 2008). 12. H. Saito, Y. Goto, Y. Mabuchi, I. Alimuddin, G. Bagtasa, N. Manago, H. Irie, I. Harada, T. Ishibashi, K. Yashiro, S. Kameyama, and H. Kuze, Open J. Air Pollut. 3, 20 (2014).
2571
13. Y. Kambe, Y. Yoshii, K. Takahashi, and K. Tonokura, J. Environ. Monit. 14, 944 (2012). 14. M. Buchwitz, R. de Beek, J. P. Burrows, H. Bovensmann, T. Warneke, J. Notholt, J. F. Meirink, A. P. H. Goede, P. Bergamaschi, S. Korner, M. Heimann, and A. Schulz, Atmos. Chem. Phys. 5, 941 (2005). 15. C. Kern, S. Trick, B. Rippel, and U. Platt, Appl. Opt. 45, 2077 (2006). 16. D. W. Marquardt, J. Soc. Ind. Appl. Math. 11, 431 (1963). 17. L. S. Rothman, I. E. Gordon, Y. Babikov, A. Barbe, D. Chris Benner, P. F. Bernath, M. Birk, L. Bizzocchi, V. Boudon, L. R. Brown, A. Campargue, K. Chance, L. Coudert, V. M. Devi, B. J. Drouin, A. Fayt, J. M. Flaud, R. R. Gamache, J. Harrison, J. M. Hartmann, C. Hill, J. T. Hodges, D. Jacquemart, A. Jolly, J. Lamouroux, R. J. LeRoy, G. Li, D. Long, C. J. Mackie, S. T. Massie, S. Mikhailenko, H. S. P. Müller, O. V. Naumenko, A. V. Nikitin, J. Orphal, V. Perevalov, A. Perrin, E. R. Polovtseva, C. Richard, M. A. H. Smith, E. Starikova, K. Sung, S. Tashkun, J. Tennyson, G. C. Toon, V. G. Tyuterev, J. Vander Auwera, and G. Wagner, J. Quant. Spectrosc. Radiat. Transfer 130, 4 (2013). 18. W. Dansgaard, Tellus 16, 436 (1964). 19. M. Berglund and M. E. Wieser, Pure Appl. Chem. 83, 397 (2011). 20. A. Mahalov and M. Moustaoui, Laser Photon. Rev. 4, 144 (2010).
