Atmospheric CO 2 retrieval from ground based FTIR spectrometer over Shadnagar , India

Abstract. The column-averaged volume mixing ratio (vmr) of carbon dioxide (XCO2) has been retrieved over Shadnagar (Latitude = 17.03° N, Longitude = 78.21° E), India at Near Infrared (NIR) spectra using ground based high resolution (∆ν = 0.01 cm−1) Fourier Transform InfraRed (FTIR; model IFS125M) spectrometer. It is first of its kind in India for measuring columnar and vertical mixing ratios of atmospheric trace gases and other greenhouse gases (GHGs). In the present study, retrieval of XCO2 could be performed for a week at near Sun's nadir view (solar zenith angle-SZA, ±75.0°) using line-by-line radiative transfer algorithm (LBLRTA). Vertical profiles of CO2 have been retrieved at NIR and middle IR (MIR) spectra using a version of the Fast Atmospheric Signature Code 3 (FASCODE3) model which is for the retrieval of atmospheric trace gas profiles. Error residuals between measured and fitted atmospheric transmission lie with in ±1.0 % for CO2 (6180–6380 cm−1) and O2 (7800–7940 cm−1) respectively. During analysis period mean (standard deviation, 1σ) XCO2 was observed to be 385.24 ppm (4.22 ppm) with 1.0 % of daily variation. Minimum and maximum averaged molecular column densities of CO2 (O2) are 6.15 × 1021 (3.3 × 1024) molecules/cm2 and 8.06 × 1021 (4.72 × 1024) molecules/cm2 respectively. Obtained an average high signal to noise ratio (SNR) of 833 and 625 for NIR and MIR spectra, respectively.


48
Greenhouse and other trace gases such as carbon dioxide (CO2), methane (CH4), ozone (O3) and 49 nitrous oxide (N2O) play a vital role in controlling the climate system of the lower atmosphere 50 (Stocker et al., 2013;Smedley et al., 2015;Sreenivas et al., 2016). CO2 in the atmosphere is the 51 most important contributor to positive radiative forcing that increases the greenhouse effect 52 (Forster et al., 2007). It has increased about 40% from the year 1750 to 2011 with the level of 53 atmospheric abundance of CO2 was 390.5 ppm (390.3 to 390.7) in the year 2011 (Stocker et al.,54 2013). To understand better and manage CO2 emissions, estimates of source and sink strengths 55 are required by the integrated approach of in-situ, remote sensing and model simulation. Currently 56 the information about atmospheric CO2 is mainly inferred from in situ (Warneke et al., 2005 convection is consistently strong and as a result flux signals are only weakly seen in surface 70 measurements (Gloor et al., 2000).  In the present study, NIR and MIR spectra have been utilized to retrieve column averaged and 81 vertical profile of CO2 using LBLRTA and radiative inverse model (FASCODE3). InSb detector to understand the typical columnar concentrations of CO2 by implementing the basic 91 LBLRTA at 6100 cm -1 -6400 cm -1 spectral range. Also attempted to retrieve CO2 vertical profile 92 information in NIR and MIR spectral range using FASCODE3 model (Notholt 1994). The 93 FASCODE3 (Smith et al., 1978;Wang et al., 1996) coupled with an inversion module is based on 94 the optimal estimation method of Rodgers (1976), provides error analysis tools necessary to 95 determine the information content of the retrievals. The reader is referred to detailed retrieval 96 methods and their error analysis explained in Miller et al. (1999). Observations covering the 97 spectral range from 6000 cm -1 to 8000 cm -1 were used to retrieve the column averaged dry mole 98 fraction (DMF) of CO2 and vmr of O2 (Wallace andLivingston, 1990 andYang et al., 2002).

99
Implemented LBLRTA for the analysis of solar spectra absorption features to retrieve the 100 columnar abundance of CO2 in the atmosphere. In the present study, the absorption bands for CO2 101 at 6100 cm -1 -6400 cm -1 and for O2 at 7800 cm -1 -7960 cm -1 have been used. Retrieval includes 102 various options such as scaling of a priori profiles of pressure (P), temperature (T) and volume 103 mixing ratios (vmr). In this study, we have used a priori profiles (vmr) simulated over Izana  The measured spectra were analyzed at various times throughout the day particularly solar zenith 116 angle (SZA) at around nadir view (±75.0°) to obtain the atmospheric signal over the study region.

117
The spectra are analyzed are generally co-additions of 2-4 individual spectra, each taking 5 minutes 118 to acquire. In our analysis, an average SZA taken to be 75°.0. Figure 2 shows a sample (21 st March 119 2016) spectral analysis in which two spectral bands of CO2 and O2 were fitted against measured.

120
FASCODE3 model has been used to fit the CO2 and O2 spectra against measured transmittance 121 spectra. It computes spectral transmittance, radiance and optical depth for a given spectral range.

122
The spectral line parameters are based on the latest HITRAN line list (Rothman et al., 2013). Heart

126
In our analysis, we used central wave numbers ( ) for CO2 and O2 are 6348 cm -1 and 7808 cm -1 127 respectively. It has an advantage of being collected using the same detector and the ratio of CO2/O2 128 will mostly cancel out the systematic effects such as instrument line shape (ILS). In figure 2b and 129 2d residual (measured-calculated) errors were shown for CO2 and O2 that lie within ±1.0%  The column-averaged CO2 was observed to be minimum (maximum) of 381.0 ppm (390.0 ppm).