- freely available
Remote Sensing 2013, 5(7), 3431-3456; doi:10.3390/rs5073431
Abstract: This paper presents a comparison of CO2 products derived from Scanning Imaging Absorption Spectrometer for Atmospheric Cartography (SCIAMACHY), Greenhouse Gases Observing Satellite (GOSAT) and Atmospheric Infrared Sounder (AIRS), with reference to calibration data obtained using the high-resolution ground-based Fourier Transform Spectrometers (g-b FTS) in the Total Carbon Column Observing Network (TCCON). Based on the monthly averages, we calculate the global offsets and regional relative precisions between satellite products and g-b FTS measurements. The results are as follows: the monthly means of SCIAMACHY data are systemically slightly lower than g-b FTS, but limited in coverage; the GOSAT data are superior in stability, but inferior in systematic error; the mean difference between AIRS data and that of g-b FTS is small; and the monthly global coverage is above 95%. Therefore, the AIRS data are better than the other two satellite products in both coverage and accuracy. We also estimate linear trends based on monthly mean data and find that the differences between the satellite products and the g-b FTS data range from 0.25 ppm (SCIAMACHY) to 1.26 ppm (AIRS). The latitudinal distributions of the zonal means of the three satellite products show similar spatial features. The seasonal cycle of satellite products also illustrates the same trend with g-b FTS observations.
Total atmospheric carbon dioxide (CO2) has increased from approximately 280 to 379 ppm over the past century, due to the burning of fossil fuels for expanding industrial activities . The annual increase in the CO2 concentration has been as high as 1.9 ppm in the decade from 1995 to 2005, which is higher than the longer term increase of 1.4 ppm per year that has been directly measured for 1960 to 2005 . CO2 absorbs infrared radiation emitted from the earth’s surface, so an increase in CO2 concentration leads to a rise in atmospheric temperatures. Temperature changes can cause feedback loops that alter CO2 concentrations by influencing the biosphere . CO2 and other greenhouse gases also influence tropospheric ozone and water vapor, further increasing their importance to the Earth’s radiative budget. Tropical land ecosystems contributed most of the interannual changes in Earth’s carbon balance through the 1980s, whereas northern mid- and high-latitude terrestrial ecosystems dominated from 1990 to 1995 . Therefore, CO2 and other greenhouse gases, such as methane (CH4), nitrous oxide (N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs) and sulfur hexafluoride (SF6), are subject to emissions regulations under the Kyoto Protocol . Research on the greenhouse effect and carbon monitoring require high precision and the long-term measurement of the atmospheric CO2 concentration. Space-based remote sensing of the CO2 column-average dry air mole fractions (XCO2) has the potential to provide observed global constraints on CO2 fluxes across the surface-atmosphere boundary and to provide insight into the related biogeochemical cycles . As satellite remote sensing technology has developed, a series of satellites that are able to detect CO2 has been launched. Satellite observations of CO2 offer new insights into the magnitude of regional sources and sinks and can help overcome the large uncertainties associated with the upscaling and interpretation of data on CO2 concentration from the Earth’s surface .
Currently, several satellite instruments can retrieve CO2 and other greenhouse gas data with significant sensitivity. These include the Scanning Imaging Absorption Spectrometer for Atmospheric Cartography (SCIAMACHY) [7,8] on board the Environmental Satellite (ENVISAT), which was launched in 2002, and the Thermal and Near-infrared Sensor for Carbon Observation (TANSO)  on board the Greenhouse Gases Observing Satellite (GOSAT), which was launched in 2009. The Atmospheric Infrared Sounder (AIRS), the Advanced Microwave Sounding Unit (AMSU) and the Humidity Sounder for Brazil (HSB) form an integrated cross-track scanning temperature and humidity sounding system on the Aqua satellite of the Earth Observing System (EOS) . The second Orbiting Carbon Observatory (OCO-2) [11,12] is another satellite designed to observe atmospheric CO2 in the same spectral region as SCIAMACHY and TANSO in the lower troposphere. The future generation of satellites, such as CarbonSat, which is to be launched in 2019 at the earliest [13–15], will also be able to constrain the parameterization of anthropogenic CO2 emissions. To study CO2 sources and sinks, China has launched the Chinese carbon dioxide observation satellite (TanSat) project [16,17]. The TanSat-Chinese will be launched in 2015 and will monitor the CO2 in the Sun-synchronous orbit.
XCO2 products retrieved from these satellites have been validated with reference to high-resolution ground-based Fourier Transform Spectrometers (g-b FTS) data and model data. Morino et al.  validated the GOSAT short wave infrared (SWIR) L2 V01.xx XCO2 products with the g-b FTS data in nine TCCON sites, and their preliminary findings were that the difference between the GOSAT XCO2 data and the g-b FTS data was −8.85 ± 4.75 ppm or −2.3 ± 1.2%. These biases and standard deviations are somewhat higher than those of other validations [5,6,18]. Yoshida et al.  identified and corrected the Version 01.xx GOSAT XCO2 products using a revised version of the retrieval algorithm (Version 02.xx). The improved retrieval algorithm had much smaller biases and standard deviations (−1.48 ppm and 2.10 ppm) for XCO2 than did Version 01.xx. Inoue et al.  used two approaches to validate the GOSAT XCO2 products (Version 02.00) with aircraft measurement data. Both methods indicated that the Version 02.00 of GOSAT XCO2 products were improvements over the previous Version, and the bias was 1–2 ppm, with a standard deviation of 1–3 ppm. Schneising et al. [21,22] compared SCIAMACHY XCO2 data to g-b FTS measurements and model results (CarbonTracker XCO2) for the period from 2003 to 2009. The relative accuracy was 1.14 ppm relative to TCCON and 1.20 ppm relative to CarbonTracker. Wang et al.  analyzed SCIAMACHY XCO2 data levels in China and found that the largest peak-to-trough amplitude of SCIAMACHY was approximately 16 ppm during 2003–2005 and that peaks occurred in the spring, with the trough in winter or autumn. Bai et al.  validated the AIRS XCO2 products from 2003 to 2008 using five ground-based stations located throughout the world. The correlation coefficients between AIRS and the other five ground stations were greater than 0.77, with a monthly mean difference of ∼0.62 ± 3.0 ppm. The average concentration of atmospheric CO2 was higher in the Northern Hemisphere than in the Southern Hemisphere, with the approximately 2 ppm per year of annual growth in China.
These XCO2 products have been validated with g-b FTS data and model data, but there are some gaps in these comparisons. These XCO2 products have been validated individually, but we do not know the differences among the XCO2 values retrieved from the three sensors (SCIAMACHY, AIRS and GOSAT). Therefore, we have no gauge of which sensors are more precise under what circumstances; additionally, we do not know the differences among these satellite products for different regions. In this paper, the satellite products of GOSAT, SCIAMACHY and AIRS are compared with the reference calibrated data obtained using g-b FTS in the TCCON sites from 2003 to 2011. Through these comparisons, we try to explain why the validations of satellite products differ, to understand the different retrieval methods of the satellite products and to give an overall evaluation. Such a comparison is a prerequisite to evaluating the precision of each product and the suitability for their use in different conditions.
The three sets of satellite data are described in Section 2, where we present an overview of these projects and compare these satellite data sets. Reference data measured with g-b FTS and comparison methods are described in Section 3. The precision analysis of the satellite data products and the preliminary comparison to the reference data are presented in Section 4, where we also analyze the spatial distribution feature, linear trend and seasonal cycle using monthly mean data. The discussion and conclusions follow.
2. Satellite Data Sets
2.1. TANSO on GOSAT
GOSAT was launched on 23 January 2009, by the Japanese Space Agency. The GOSAT Project is a joint effort of the Ministry of the Environment (MOE), the National Institute for Environmental Studies (NIES) and the Japan Aerospace Exploration Agency (JAXA) with the investment of 20 billion Japanese yen . The GOSAT Project primarily estimates the emission and absorption of greenhouse gases on a subcontinental scale. It flies at a 666 km altitude and completes an orbit in approximately 100 min, with an equator crossing time of approximately 1:00 a.m. local time. Its mission is to provide global measurements of total column XCO2 and XCH4 from the short wave infrared bands, and it returns to observe the same point on Earth every three days. The instrument on board the satellite is the Thermal and Near-infrared Sensor for Carbon Observation (TANSO), which is composed of two subunits: the Fourier Transform Spectrometer (FTS) and the Cloud and Aerosol Imager (CAI).
Carbon observation is made by an FTS that covers 0.75∼14.3 μm : band 1 spans 0.758–0.775 μm (12,900–13,200 cm−1) with 0.37 cm−1 or better spectral resolution, and bands 2–4 span 1.56–1.72, 1.92–2.08 and 5.56–14.3 μm (5,800–6,400 cm−1, 4,800–5,200 cm−1 and 700–1,800 cm−1) respectively, with 0.26 cm−1 or better spectral resolution. The TANSO instantaneous field of view is approximately 15.8 millradians, corresponding to a nadir footprint diameter of approximately 10.5 km at sea level. The nominal single-scan data acquisition time is 4 s.
XCO2 and XCH4 are retrieved from the 1.6 μm CO2 absorption band and the 1.67 μm CH4 absorption band using the optimal estimation method . The retrieval algorithm for XCO2 and XCH4 is described in Yoshida . It consists of three parts: screening data suitable for the retrieval analyses, optimal estimation of gaseous column abundances and examination of the retrieval quality to exclude low-quality and/or aerosol-contaminated results. Most of the random errors in retrieval come from instrumental noise, while the interference error due to auxiliary parameters is relatively small. GOSAT Level 2 products are evaluated against high-precision data that are obtained independently using ground-based or aircraft observations. GOSAT Level 3 products are generated by interpolating and extrapolating the Level 2 products and can be used to estimate the global distribution of greenhouse gas concentrations. The Level 2 products indicate the analytical value of the amount of greenhouse gas observed at a specific observation point at a specific observation time, and Level 3 products provide a monthly global distribution of greenhouse gases.
2.2. SCIAMACHY on ENVISAT
SCIAMACHY, on board the European environmental satellite, was launched into a Sun-synchronous orbit in a descending node and has an equator crossing time of 10:00 a.m. local time. Its wavelength coverage is from 240 nm to 2,380 nm, with eight observation channels that measure reflected, backscattered and transmitted solar radiation at moderately high spectral resolution (0.2–1.4 nm) in the spectral region [29,30]. The Weighting Function Modified Differential Optical Absorption Spectroscopy (WFM-DOAS) algorithm [31,32], which was produced by the University of Bremen, is an important improvement to the standard DOAS algorithm. The global long-term SCIAMACHY greenhouse gas results for validation are obtained using v2 (v2.1 XCO2) of the scientific retrieval algorithm, WFM-DOAS, that was introduced by Schneising et al. , which is based on a fast look-up table scheme. WFM-DOAS is a least-squares method based on scaling (or shifting) pre-selected atmospheric vertical profiles. XCO2 is derived by normalizing the CO2 columns with the simultaneously retrieved oxygen columns from the O2 A-band. The monthly average concentration data for XCO2 from January 2003 to December 2009 were retrieved from SCIAMACHY nadir observation mode L1c data combined with the WFM-DOAS algorithm.
The overall precision and bias of the XCO2 are estimated to be close 1.0% and < 4.0% . The averaging daily standard deviations of the retrieved XCO2 at different locations are approximately 5.4 ppm (1.4%). Reuter et al.  validated the XCO2 retrieved from Bremen Optimal Estimation DOAS algorithm by comparing them with g-b FTS measurements and with CT2009 from 2006 to 2010. The single-measurement precision is no more than 3 ppm, and the year-to-year increase varies between 1.88 ± 0.44 and 2.30 ± 0.03 ppm per year.
2.3. AIRS on Aqua
AIRS is designed to meet the requirements of the NASA Earth Science Enterprise climate research programs and the operational weather-forecasting plans of the National Oceanic and Atmospheric Administration. It is a nadir cross-track scanning infrared spectrometer on the second Earth Observing System  Aqua spacecraft, which flies at an altitude of approximately 705 km polar orbit. The sensor on board the satellite has 2,378 channels that cover three spectral regions from 649 to 2,674 cm−1 (649–1,136, 1,217–1,613 and 2,169–2,674 cm−1). It crosses the equator at approximately 1:30 a.m. and 1:30 p.m. local time, resulting in near global coverage twice a day . The instrument field of view is 1.1°, corresponding to a nadir footprint of 13.5 km on the surface, and the scan angles are ±48.95° . AIRS can produce cloud cleared radiance for approximately 60% of the 324,000 FORs (field of regards) per day.
The Version 5 Level 2 retrieval algorithm (V5) that is used to retrieve these products assumes a global average linear time-varying CO2 climatology throughout the atmosphere . The CO2 transport model is an atmospheric four-dimensional variational data assimilation system, which is a practical formulation of Bayesian estimation theory . AIRS products meet the criteria identified by the National Research Council for climate data records. XCO2 retrieved from AIRS uses two thermal infrared strong absorption bands, which correspond to low and mid-high sensitivities to tropospheric concentration.
2.4. Comparisons among Three Satellite Data Sets
For long-term comparison with g-b FTS data, we use the monthly mean data from the satellite products. The global monthly average XCO2 data for GOSAT (L3 SWIR v01.xx) come from the NIES GOSAT website ( http://www.gosat.nies.go.jp/index_e.html) from the period of April 2009 to December 2011, with a sampling grid of 2.5° × 2.5°. The global monthly average XCO2 data for SCIAMACHY (L3 WFM-DOAS v2.2) come from the SCIAMACHY/WFM-DOAS research team at Germany’s University of Bremen ( http://www.iup.uni-bremen.de/sciamachy/NIR_NADIR_WFM_DOAS/) for the period of January 2003 to December 2009, with a sampling grid of 0.5° × 0.5°. The global monthly average XCO2 data for AIRS (L3 AIRS+AMSU v5) come from the NASA AIRS website ( http://disc.sci.gsfc.nasa.gov/AIRS/data-holdings) in the period of January 2002 to December 2011, with a sampling grid of 2° × 2.5°.
The comparison of the performance of the satellite sensors for the three on-orbit satellites, GOSAT, SCIAMACHY and AIRS, are listed in Table 1. There are two salient features of the various CO2 observation parameters: (1) a wide band range with high spectral resolution, but relatively lower spatial resolution; and (2) nadir observation mode in all sensors, except SCIAMACHY, which uses a variety of observation mode. The boundary layer observation mode of SCIAMACHY has higher vertical resolution and greater sensitivity, and can obtain the higher precision atmospheric vertical profile of the troposphere and stratosphere . Some characteristics of the CO2 absorption band and other necessary spectral characteristic channel of retrieval parameters are all include in the spectral range. AIRS mainly uses the thermal infrared absorption band for CO2 near 4.3 μm and 1.5 μm, which is most sensitive to CO2 in the low and mid-high layers. SCIAMACHY primarily uses the short-wave infrared weak absorption band of CO2 near 1.58 μm, which is most sensitive to CO2 near the ground. In addition to these, GOSAT uses the short wave infrared weak absorption band of CO2 near 2.06 μm, which covers the most general CO2 characteristic bands in the current monitoring of CO2 and has great significance for CO2 detection band selection in future platforms .
Figure 1 shows the global XCO2 distribution of the monthly averaged AIRS, GOSAT and SCIAMACHY, gridded in 2° by 2.5° bins, 2.5° by 2.5° bins and 0.5° by 0.5° in April and October, 2009, respectively. For AIRS, data covering almost the entire surface of the Earth are obtained for 60°S to 80°N in both April and October. For GOSAT, terrestrial data are obtained mainly for 10–60°N and 15–45°S in April and for 10–50°N and 0–50°S in October. Data over the oceans are retrieved from 10°S to 30°N in April and from 40°S to 10°N in October by observing the specular reflection of sunlight in the direction of sunlight.
In Figure 1, we can see that the SCIAMACHY data are confined to the land. The GOSAT data cover the ocean and some nearshore sea areas within the latitudes of 40°S–40°N. The coverage of AIRS data has greatly improved; with the exception of latitudes to the south of 60°S near the arctic pole point, the monitoring scope of AIRS covers most of the world. The global data coverage for a single month for the three satellites varies: the SCIAMACHY products cover only 8–20%, GOSAT products improve that figure to 32–44% and AIRS products cover 95%, which provides an obvious advantage compared to the other types of satellite data.
The spatial differences in XCO2 between three sensor measurements are mainly due to the product selection for validation. The GOSAT XCO2 data products shown here are filtered for aerosol optical depth less than 0.5. As a plane-parallel atmosphere is assumed in the retrieval, data with solar zenith angles greater than 70 degrees are not processed, and data over high mountain ranges, such as the Rockies, the Andes and the Himalayan mountains, are removed . SCIAMACHY measurements over the ocean are filtered out, because of the lower surface reflectance. In order to get larger global data coverage scale, the AIRS XCO2 product lost some accuracy in the product process. Thermal infrared band sensitive channels used in the AIRS XCO2 product are superior to the near-infrared wave band for the ocean detection ability of CO2. Therefore, AIRS has stronger detection ability for CO2 over the ocean compared with the other kinds of satellites.
3. Reference Data and Comparison Methodology
3.1. g-b FTS Data for Satellite Product Validation
TCCON is a global network of ground-based high resolution FTS that record direct solar spectra in the near-infrared spectral region . The g-b FTS data are calibrated using airborne in situ measurements and apply single scaling factors for each species to all sites consistently. The precision of g-b FTS measurements of XCO2 are better than 0.2% under clear skies [42–45]. They provide validation data for column-averaged satellite data with accurate and precise column-averaged abundances of CO2 and CH4.
To ensure data comparability, all TCCON sites use similar instrumentation (Bruker IFS 125/HR for all sites in this study) and a common retrieval algorithm based on scaling a priori profiles by least-squares fitting . The CO2a priori profiles are derived from an empirical model based on GLOBALVIEW in situ data and extended to the stratosphere using an age of air relationship .
The comparison was performed at the following TCCON sites: Białystok (Poland), Bremen (Germany), Garmisch (Germany), Park Falls (USA), Lamont (USA), Tsukuba (Japan), Izaña (Spain), Darwin (Australia), Wollongong (Australia) and Lauder (New Zealand). The g-b FTS data from the ten TCCON sites are used in this analysis. Table 2 summarizes the spatial coordinates of those stations. Figure 2 shows the locations of the TCCON sites that are used in this paper.
For each TCCON site, an XCO2 time series from SCIAMACHY, GOSAT, AIRS and g-b FTS was generated for validation and comparison. AIRS data are available for the entire analyzed time period ranging from 2003 to 2011, whereas SCIAMACHY and GOSAT data are available for certain subperiods, i.e., 2003 to 2009 and 2009 to 2011, respectively. Similarly, g-b FTS data are also available for certain subperiods depending on the site (Park Falls and Lauder since 2004, Darwin since 2005, Bremen and Izaña since 2007, Tsukuba, Lamont and Wollongong since 2008 and Białystok and Garmisch since 2009). However, in Izaña, there are too few WFM-DOAS retrievals that passed the quality filter for a statistically significant comparison, so no SCIAMACHY data are available for the analyzed time period. Therefore, the comparison between satellite products and g-b FTS data is presented not only for the entire analyzed time period, but also for two subperiods, from January 2003 to December 2008, and January 2009 to December 2011. Thereby, we can see whether these products are stable in different periods of performance.
3.2. Comparison Methodology
The monthly mean values of the g-b FTS data are calculated from the total of the daily TCCON site measurements divided by the number of days of the month. Additionally, no corrections (on the a priori or on the averaging kernel) have been considered in the manuscript. The time series of the satellite product’s individual mean volume mixing ratio (vmr) in the ith TCCON site is defined as , and the corresponding monthly mean g-b FTS value is defined as . Accordingly, the relative difference between the selected satellite products and the g-b FTS data is:
The mean difference between satellite products and g-b FTS data can be calculated using the following formula:
4.1. Regional Precision and Relative Accuracy
The XCO2 results for each satellite product are listed in Table 3 for each site, showing the mean differences (d) to g-b FTS, the standard deviations of the differences (s) and the correlation coefficients (r) for the analyzed time period and the two subperiods. Table 4 summarizes the results. The two subperiods are January 2003 to December 2008, and January 2009, to December 2011. The corresponding monthly time series and relative differences are depicted in Figures 3 and 4 for the entire analyzed time period for Bialystok, Bremen, Garmisch, Park Falls, Lamont, Tsukuba, Izaña, Darwin, Wollongong and Lauder.
Another more important error estimate is the relative accuracy: the relative accuracy of GOSAT data amounts to 1.50 ppm (0.38%) relative to g-b FTS data, that of SCIAMACHY is 3.35 ppm (0.87%) and that of AIRS is 0.96 ppm (0.25%) for the entire analyzed time period. Overall, we find a good agreement between the AIRS data and the reference data, and the GOSAT data present a good agreement with the g-b FTS data in the subperiod from 2009 to 2011.
Scatter diagrams of the time series of the satellite products and g-b FTS data for XCO2 with the correlation coefficients at selected TCCON sites are shown in Figure 5. We only plotted data for which the g-b FTS data were averaged for a month and the corresponding satellite XCO2 values were successfully retrieved. Therefore, there are fewer satellite data points in the scatter diagrams than are listed in the time series. The g-b FTS data at Darwin were not obtained between February, 2010, and August, 2010, due to mechanical problems with the sun tracker. Only a few GOSAT and SCIAMACHY data are available for comparison with Bialystok, Bremen, Gamisch, Tsukuba and Lauder, and no SCIAMACHY data are available at the Izaña site. SCIAMACHY data have an opposing trend from the g-b FTS data at Garmisch, mainly because of fewer available data. Due to using the monthly mean data, the mean correlation coefficients of the three satellite data are less than 0.5.
4.2. Seasonal Accuracy
To obtain the seasonal accuracy of all types of satellite products, we performed a seasonal comparison for them. The comparison methodology is similar to Section 3.2, but the time interval is different. The entire analyzed time period can be divided into four subperiods according to four seasons. The results of seasonal comparisons for each site are shown in Table 5. Table 6 summarizes the results.
The error estimate for each satellite product in the four seasons is the relative accuracy listed in Table 6. In spring, the global offset and relative accuracy of AIRS data are as low as 1.15 ppm and 1.11 ppm (0.30%), respectively. In summer, the relative precision and accuracy of GOSAT data are 1.43 ppm (0.09%) and 2.31 ppm (2.23%), respectively. In autumn, the global offset of SCIAMACHY data is only 1.16 ppm, but the regional bias is a little higher at 4.03 ppm (1.06%). In winter, the comparative data sets are fewer, except for AIRS, which has global offset and regional biases of 0.65 ppm and 1.30 ppm (0.46%), respectively. Because the AIRS data are averaged within a larger scope, the fluctuation range is relatively smaller than 1.3 ppm.
4.3. Annual Increase and Seasonal Cycle
We calculated the linear trends of the year separately in cases where data for the year are available for a sufficient number (more than three years of data). First, we calculate the increment of the ith month in a year as follows:
Therefore, we can obtain up to 12 individual trends (e.g., one for all Januarys). We can then calculate the linear trend of each site, , and the annual increase is the mean of the trend, , where n is the number of increments. The annual increases listed in Table 7 are derived by fitting a linear trend to the deseasonalized time series. The mean amplitude of the seasonal cycle is obtained by subtracting the linear trend derived above from the time series. Then, we average the resulting amplitudes for all years in which the cycle is reasonably sampled, denoting the standard error of the mean as an error. The average peak-to-peak amplitude of the seasonal cycles is also listed in Table 7; std is the standard deviation of the increments. Time series of the XCO2 yearly linear increasing trends are shown in Figure 6.
According to Table 7, there are generally no significant differences between satellite data and the annual increase according to g-b FTS data. The Lauder site has a seasonal cycle in XCO2 with a small peak-to-peak amplitude of approximately 0.6 ppm. The measurements over Wollongong reach approximately 2 ppm peak-to-peak. The Lauder XCO2 time series is the longest in the Southern Hemisphere and has had a consistent increase of 1.89 ppm per year since 2004 . For all sites, the increases agree within their errors, with the exception of Park Falls and Lamont, where a marginal residual difference remains after considering the standard errors for SCIAMACHY . Nevertheless, the global mean increase is somewhat larger for satellite data than for g-b FTS data (1.78 ± 1.68, 1.06 ± 5.43 and 2.07 ± 1.62 ppm per year compared to 0.81 ± 1.12 ppm per year). The highest annual increase among the three satellite products is measured by AIRS, at 2.07 ppm per year. However, the TCCON increase at Bremen is significantly larger than the one for the GOSAT and SCIAMACHY time series (1.81 ± 2.23 ppm per year compared to 1.76 ± 2.07 and 0.24 ± 4.81 ppm per year).
To obtain a quantitative impression of the global seasonal cycle differences and their significance, we calculate the mean amplitudes for all TCCON sites where estimated seasonal cycle amplitudes are available for all three data sets. After subtracting the linear trend, i.e., the year-to-year increase, we calculate the average peak-to-peak amplitude of the seasonal cycles. The seasonal cycles in all data sets are in close agreement. The average peak to peak amplitudes are 1.37 ± 0.61 ppm for the g-b FTS, 2.21 ± 0.92 ppm for GOSAT, 6.39 ± 4.14 ppm for SCIAMACHY and 2.75 ± 1.32 ppm for AIRS. Due to SCIAMACHY’s relatively large standard error, there are significant differences between SCIAMACHY and the other data sets. The amplitudes derived from SCIAMACHY are typically larger than those from other satellite data sets, which are, in turn, larger than those from TCCON.
4.4. Latitudinal Distribution of Zonal Averaged Satellite Data
To analyze the spatial consistency and regional differences between remote sensing data, the monthly means of zonal averaged satellite data are calculated for each 10° band of latitude. Latitudinal distributions of monthly means of zonal averaged satellite data of XCO2 in April 2009, and in October, 2009, are shown in Figure 7. In summer and autumn, the strong photosynthesis by lush vegetation leads to the decrease of the CO2 concentration until its minimum in October. In winter and spring, the weak photosynthesis of the withered vegetation and winter heating lead to higher CO2 concentrations, which reach peak in April.
All data sets show that XCO2 is higher in the Northern Hemisphere compared to the Southern Hemisphere in April, but the difference between the hemispheres is small in October. The GOSAT data are approximately 10 ppm lower in the northern mid-latitudes and nearly 15 ppm lower in the southern mid-latitudes than the other data sets.
Column average dry air mole fractions of CO2 from 2003 to 2011 have been compared using GOSAT (2009–2011), SCIAMACHY (2003–2009), AIRS (2003–2011) and g-b FTS measurements. The comparison is also performed for global zonal averages latitudinal from 60° S to 80° N in 10° steps. The comparison is performed by comparing monthly mean data in the entire analyzed time period and two subperiods, respectively. Both approaches yield similar results.
Among the three satellite products, the accuracy of SCIAMACHY data is relatively low, and the observation range is limited. The scatter of GOSAT retrievals is well below 1%, though this is a substantial improvement over earlier GOSAT validation efforts [4,28]. The XCO2 data retrieved from SCIAMACHY perform better in inland areas. The values of SCIAMACHY are lower than g-b FTS, but the difference is never more than 2 ppm. The XCO2 data retrieved from SCIAMACHY are inferior in offshore areas and the difference is larger. The XCO2 data values retrieved from GOSAT are lower than g-b FTS and the difference amounts to 7 ppm. However, the GOSAT data do match well with g-b FTS on islands. This finding shows that the stability and correlation of GOSAT data have improved globally, especially in oceanic and offshore areas. The AIRS data products have the distinct advantage in terms of coverage range and measurement accuracy, which also provides more opportunity for comparisons with g-b FTS data. The absence of some high values in the AIRS data is consistent with the mid- and upper troposphere, being less sensitive to surface emissions . We also sampled both SCIAMACHY and AIRS data sets within ±2.5 degrees and, then, compared with the original data. The results are listed in Table 8.
The differences of global offset and mean correlation are small. However, the different spatial resolution imperfect match has significantly contributed to the observed standard deviations .
We have also estimated linear trends from monthly mean data for different periods. In general, the XCO2 seasonal cycle amplitudes derived from satellite data are somewhat larger than those from g-b FTS data. Because the time periods considered at certain TCCON sites are rather short, these trend estimates are only used to identify potential instrumental issues with the GOSAT and SCIAMACHY data. Although the XCO2 retrievals from AIRS are in the troposphere, the products are more consistent with the g-b FTS measurements.
The relative accuracy of AIRS data is 1.11 ppm (0.30%) and 1.08 ppm (0.28%) in spring and autumn, respectively. The relative accuracy of GOSAT data is 2.31 ppm (2.23%) in summer and SCIAMACHY data is 1.68 ppm (0.34%) in winter. Although there are different biases for XCO2, the satellite data retrievals and g-b FTS data show similar seasonal behaviors over the Northern Hemisphere: higher in spring and lower in autumn. The seasonal change in atmospheric CO2 concentration is mainly influenced by CO2 absorption and emission by terrestrial ecosystems. In summer and autumn, the strong photosynthesis of the lush vegetation leads to the decrease of the CO2 concentration and reaches to a low in October. In winter and spring, weak photosynthesis and winter heating lead to higher CO2 concentrations, which maintain a higher level and peak in April.
In this paper, we analyze the XCO2 products retrieved from Imaging Absorption Spectrometer for Atmospheric Cartography (SCIAMACHY), Greenhouse Gases Observing Satellite (GOSAT), Atmospheric Infrared Sounder (AIRS) and use high-resolution ground-based Fourier Transform Spectrometers (g-b FTS) with reference to calibration data obtained at ten Total Carbon Column Observing Network (TCCON) from 2003 to 2011. The validation and comparison reveal that the satellite data are biased by −6.53 ± 1.77 ppm (0.46%), −0.78 ± 4.67 ppm (1.22%) and 1.84 ± 2.76 ppm (0.72%) for GOSAT, SCIAMACHY and AIRS relative to the reference values, respectively. The relative accuracy of these satellite data are 1.50 ppm (0.38%), 3.35 ppm (0.87%) and 0.96 ppm (0.25%), respectively, and the mean correlation coefficients are 0.64, 0.43 and 0.56 compared to g-b FTS data, respectively. The mean annual increase in XCO2 of GOSAT, SCIAMACHY, AIRS and g-b FTS data are 1.78, 1.06, 2.07 and 0.81 ppm per year, respectively. The relative accuracy of AIRS data is 1.11 ppm (0.30%) and 1.08 ppm (0.28%) in spring and autumn, respectively. The relative accuracy of GOSAT data is 2.31 ppm (2.23%) in summer and SCIAMACHY data, 1.68 ppm (0.34%) in winter. Due to the product selection for validation, the spatial distributions in XCO2 between three sensor measurements are different. The SCIAMACHY data are confined to the land, because of lower surface reflectance; GOSAT XCO2 data are filtered for aerosol optical depth less than 0.5 and are removed over high mountain ranges. AIRS XCO2 product get larger global data coverage scale by losing accuracy in the product process.
GOSAT data offer certain advantages, but also have high systemic bias. The XCO2 data retrieved from SCIAMACHY perform better in inland areas, but the difference is larger. Compared to the differences between the former two satellites, the system bias of AIRS is lowest for each site. Especially in mid-latitude regions, the AIRS XCO2 values agree very well with g-b FTS. Overall, AIRS data show largely the same trend as the g-b FTS data. Therefore, the AIRS data product can reflect the distribution characteristics and change rules for global CO2 atmospheric concentration very well. The uncertainties in sensor measurements are about the cloud, lower surface reflectance and systematic deviations, due to aerosols, thin cirrus clouds and even spectrum instrument depreciation of sensor. The accuracies of satellite data is also affected by the accuracies of retrieval algorithm. In the future, we plan to investigate interferences by aerosols and thin cirrus clouds using aerosol LIDaRs and/or sky-radiometers at selected FTS sites. Further improvements could be achieved by evaluating systematic errors with an empirical neural-network-type multivariate regression with physically meaningful regression coefficients in an approach similar to  or by accounting for additional physical parameters in future models (e.g., cloud parameters, as in [34,53]).
We thank the members of the GOSAT Project (JAXA, NIES and Ministry of the Environment, Japan) for providing GOSAT Level 2 data products. We thank the University of Bremen for providing us with the SCIAMACHY Level 2 data products. We thank NASA Goddard Earth Science Data and Information Services Center for providing us with the AIRS Version 5 Level 3 CO2 data products. We also thank NASA’s Terrestrial Ecology Program and the Orbiting Carbon Observatory for their support of TCCON.
This study was supported by the Public Welfare Special Program, Ministry of Environmental Protection of the People’s Republic of China (grant 200909018), the National Natural Science Foundation of China (No. 41001215), the Youth Science Funds of LREIS, CAS., and the Gray Haze Remote Sensing Monitoring Technology Research and Operation Demonstration of JiangSu Province (Grant No. 201130).
Conflict of Interest
The authors declare no conflict of interest.
References and Notes
- Intergovernmental Panel on Climate Change (IPCC). Climate Change 2007: The Physical Science Basis: Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change; Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L., Eds.; Cambridge University Press: Cambridge, UK and New York, NY, USA, 2007; p. 996. [Google Scholar]
- Cox, P.M.; Betts, R.A.; Jones, C.D.; Spall, S.A.; Totterdell, I.J. Acceleration of global warming due to carbon-cycle feedbacks in a coupled climate model. Nature 2000, 408, 184–187. [Google Scholar]
- Bousquet, P.; Peylin, P.; Ciais, P.; Le Quéré, C.; Friedlingstein, P.; Tans, P.P. Regional changes in carbon dioxide fluxes of land and oceans since 1980. Science 2000, 290, 1342–1346. [Google Scholar]
- Morino, I.; Uchino, O.; Inoue, M.; Yoshida, Y.; Yokota, T.; Wennberg, P.O.; Rettinger, M. Preliminary validation of column-averaged volume mixing ratios of carbon dioxide and methane retrieved from GOSAT short-wavelength infrared spectra. Atmos. Meas. Tech 2011, 4, 1061–1076. [Google Scholar]
- Butz, A.; Guerlet, S.; Hasekamp, O.; Schepers, D.; Galli, A.; Aben, I.; Warneke, T. Toward accurate CO2 and CH4 observations from GOSAT. Geophys. Res. Lett 2011, 38, L14812. [Google Scholar]
- Parker, R.; Boesch, H.; Cogan, A.; Fraser, A.; Feng, L.; Palmer, P. I.; Wunch, D. Methane observations from the Greenhouse Gases Observing SATellite: Comparison to ground-based TCCON data and model calculations. Geophys. Res. Lett 2011, 38, L15807. [Google Scholar]
- Burrows, J.P.; Hölzle, E.; Goede, A.P.H.; Visser, H.; Fricke, W. SCIAMACHY-Scanning imaging absorption spectrometer for atmospheric chartography. Acta Astronaut 1995, 35, 445–451. [Google Scholar]
- Bovensmann, H.; Burrows, J.P.; Buchwitz, M.; Frerick, J.; Noël, S.; Rozanov, V.V.; Goede, A.P.H. SCIAMACHY: Mission objectives and measurement modes. J. Atmos. Sci 1999, 56, 127–150. [Google Scholar]
- Yokota, T.; Yoshida, Y.; Eguchi, N.; Ota, Y.; Tanaka, T.; Watanabe, H.; Maksyutov, S. Global concentrations of CO2 and CH4 retrieved from GOSAT: First preliminary results. Sola 2009, 5, 160–163. [Google Scholar]
- Aumann, H.H.; Chahine, M.T.; Gautier, C.; Goldberg, M.D.; Kalnay, E.; McMillin, L.M.; Susskind, J. AIRS/AMSU/HSB on the Aqua mission: Design, science objectives, data products, and processing systems. IEEE Trans. Geosci. Remote Sens 2003, 41, 253–264. [Google Scholar]
- Crisp, D.; Atlas, R.M.; Breon, F.M.; Brown, L.R.; Burrows, J.P.; Ciais, P.; Schroll, S. The orbiting carbon observatory (OCO) mission. Adv. Space. Res 2004, 34, 700–709. [Google Scholar]
- Boesch, H.; Baker, D.; Connor, B.; Crisp, D.; Miller, C. Global characterization of CO2 column retrievals from shortwave-infrared satellite observations of the Orbiting Carbon Observatory-2 mission. Remote Sens 2011, 3, 270–304. [Google Scholar]
- Bovensmann, H.; Buchwitz, M.; Burrows, J. P.; Reuter, M.; Krings, T.; Gerilowski, K.; Erzinger, J. A remote sensing technique for global monitoring of power plant CO2 emissions from space and related applications. Atmos. Meas. Tech 2010, 3, 781–811. [Google Scholar]
- Krings, T.; Gerilowski, K.; Buchwitz, M.; Reuter, M.; Tretner, A.; Erzinger, J.; Bovensmann, H. MAMAP—A new spectrometer system for column-averaged methane and carbon dioxide observations from aircraft: retrieval algorithm and first inversions for point source emission rates. Atmos. Meas. Tech 2011, 4, 1735–1758. [Google Scholar]
- Velazco, V.; Buchwitz, M.; Bovensmann, H.; Reuter, M.; Schneising, O.; Heymann, J.P.; Burrows, J.P. Towards space based verification of CO2 emissions from strong localized sources: fossil fuel power plant emissions as seen by a CarbonSat constellation. Atmos. Meas. Tech 2011, 4, 2809–2822. [Google Scholar]
- Liu, Y.; Yin, Z.; Yang, Z.; Zheng, Y.; Yan, C.; Yang, D. Chinese Carbon Dioxide Observation Satellite (TanSat) Project. Proceedings of AGU Fall Meeting, San Francisco, CA, USA, 3–7 December 2011. Abstract #A42D-02..
- Liu, Y.; Yang, D.; Cai, Z. A retrieval algorithm for TanSat XCO2 observation: Retrieval experiments using GOSAT data. Chinese Sci. Bull 2013, 58, 1–4. [Google Scholar]
- Cogan, A.J.; Boesch, H.; Parker, R.J.; Feng, L.; Palmer, P.I.; Blavier, J.F.; Wunch, D. Atmospheric carbon dioxide retrieved from the Greenhouse gases Observing SATellite (GOSAT): Comparison with ground-based TCCON observations and GEOS-Chem model calculations. J. Geophys. Res 2012, 117, D21301, doi:10.1029/2012JD018087, 1–17.. [Google Scholar]
- Yoshida, Y.; Kikuchi, N.; Morino, I.; Uchino, O.; Oshchepkov, S.; Bril, A.; Yokota, T. Improvement of the retrieval algorithm for GOSAT SWIR XCO2 and XCH4 and their validation using TCCON data. Atmos. Meas. Tech 2013, 6, 949–988. [Google Scholar]
- Inoue, M.; Morino, I.; Uchino, O.; Miyamoto, Y.; Yoshida, Y.; Yokota, T.; Patra, P.K. Validation of XCO2 derived from SWIR spectra of GOSAT TANSO-FTS with aircraft measurement data. Atmos. Chem. Phys. Discuss 2013, 13, 3203–3246. [Google Scholar]
- Schneising, O.; Buchwitz, M.; Reuter, M.; Heymann, J.; Bovensmann, H.; Burrows, J.P. Long-term analysis of carbon dioxide and methane column-averaged mole fractions retrieved from SCIAMACHY. Atmos. Chem. Phys 2011, 11, 2863–2880. [Google Scholar]
- Schneising, O.; Bovensmann, H.; Buchwitz, M.; Burrows, J.P.; Deutscher, N.M.; Griffith, D.W.; Wunch, D. Atmospheric greenhouse gases retrieved from SCIAMACHY: Comparison to ground-based FTS measurements and model results. Atmos. Chem. Phys 2012, 12, 1527–1540. [Google Scholar]
- Wang, K.; Jiang, H.; Zhang, X.; Zhou, G. Analysis of spatial and temporal variations of carbon dioxide over China using SCIAMACHY satellite observations during 2003–2005. Int. J. Remote Sens 2011, 32, 815–832. [Google Scholar]
- Bai, W.; Zhang, X.; Zhang, P. Temporal and spatial distribution of tropospheric CO2 over China based on satellite observations. Chinese Sci. Bull 2010, 55, 3612–3618. [Google Scholar]
- Kuze, A.; Suto, H.; Nakajima, M.; Hamazaki, T. Thermal and near infrared sensor for carbon observation Fourier-transform spectrometer on the Greenhouse Gases Observing Satellite for greenhouse gases monitoring. Appl. Opt 2009, 48, 6716–6733. [Google Scholar]
- Hamazaki, T.; Kaneko, Y.; Kuze, A.; Kondo, K. Fourier transform spectrometer for Greenhouse Gases Observing Satellite (GOSAT). Proc. SPIE 2004, 5659, 73–80. [Google Scholar]
- Rodgers, C.D. Optimal Linear Inverse Methods. In Inverse Methods for Atmospheric Sounding: Theory and Practice; World Scientific: Singapore, 2000; pp. 65–79. [Google Scholar]
- Yoshida, Y.; Ota, Y.; Eguchi, N.; Kikuchi, N.; Nobuta, K.; Tran, H.; Yokota, T. Retrieval algorithm for CO2 and CH4 column abundances from short-wavelength infrared spectral observations by the Greenhouse Gases Observing Satellite. Atmos. Meas. Tech 2011, 4, 717–734. [Google Scholar]
- Burrows, J.P.; Schneider, W.; Geary, J.C.; Chance, K.V.; Goede, A.P.H.; Aarts, H.J.M.; de Vries, J.; Smorenburg, C.; Visser, H. Atmospheric Remote Sensing with SCIAMACHY. Processing of Digest of Topical Meeting on Optical Remote Sensing of the Atmosphere, Washington, DC, USA, 1990; pp. 71–74.
- Burrows, J.P.; Chance, K.V. Scanning imaging absorption spectrometer for atmospheric chartography. Proc. SPIE 1991, 1490, 146–155. [Google Scholar]
- Buchwitz, M.; De Beek, R.; Noël, S.; Burrows, J.P.; Bovensmann, H.; Bremer, H.; Heimann, M. Carbon monoxide, methane and carbon dioxide columns retrieved from SCIAMACHY by WFM-DOAS: year 2003 initial data set. Atmos. Chem. Phys 2005, 5, 3313–3329. [Google Scholar]
- Weber, M.; Lamsal, L.N.; Burrows, J.P. Improved SCIAMACHY WFDOAS Total Ozone Retrieval: Steps towards Homogenising Long-Term Total Ozone Datasets from GOME, SCIAMACHY, and GOME2. Processing of Envisat Symposium 2007, Montreux, Switzerland, 23–27 April 2007. ESA SP–636..
- Barkley, M.P.; Monks, P.S.; Friess, U.; Mittermeier, R.L.; Fast, H.; Körner, S.; Heimann, M. Comparisons between SCIAMACHY atmospheric CO2 retrieved using (FSI) WFM-DOAS to ground based FTIR data and the TM3 chemistry transport model. Atmos. Chem. Phys 2006, 6, 4483–4498. [Google Scholar]
- Reuter, M.; Bovensmann, H.; Buchwitz, M.; Burrows, J.P.; Connor, B.J.; Deutscher, N.M.; Wunch, D. Retrieval of atmospheric CO2 with enhanced accuracy and precision from SCIAMACHY: Validation with FTS measurements and comparison with model results. J. Geophys. Res 2011, 116, D04301. [Google Scholar]
- Pagano, T.S.; Aumann, H.H.; Hagan, D.E.; Overoye, K. Prelaunch and in-flight radiometric calibration of the Atmospheric Infrared Sounder (AIRS). IEEE Trans. Geosci. Remote Sens 2003, 41, 265–273. [Google Scholar]
- Xiong, X.; Barnet, C.; Maddy, E.; Sweeney, C.; Liu, X.; Zhou, L.; Goldberg, M. Characterization and validation of methane products from the Atmospheric Infrared Sounder (AIRS). J. Geophys. Res 2008, 113, G00A01, doi:10.1029/2007JG000500, 1–14.. [Google Scholar]
- Maddy, E.S.; Barnet, C.D.; Goldberg, M.; Sweeney, C.; Liu, X. CO2 retrievals from the Atmospheric Infrared Sounder: Methodology and validation. J. Geophys. Res 2008, 113, D11301. [Google Scholar]
- Engelen, R.J.; Serrar, S.; Chevallier, F. Four-dimensional data assimilation of atmospheric CO2 using AIRS observations. J. Geophys. Res 2009, 114, D03303, doi:10.1029/2008JD010739, 1–12.. [Google Scholar]
- He, Q.; Yu, T.; Cheng, T.H.; Gu, X.F.; Xie, D.H.; Wu, Y. Atmospheric carbon dioxide satellite remote sensing retrieval accuracy inspection and spatio-temporal characteristics analysis. J. Geo-Info. Sci 2012, 14, 250–257. [Google Scholar]
- Shi, G.Y.; Dai, T.; Xu, N. Latest progress of the study of atmospheric CO2 concentration retrievals from satellite. Adv. Earth Sci 2010, 25, 7–13. [Google Scholar]
- Wunch, D.; Toon, G.C.; Blavier, J.F.L.; Washenfelder, R.A.; Notholt, J.; Connor, B.J.; Wennberg, P.O. The total carbon column observing network. Philos. T. R. Soc. A 2011, 369, 2087–2112. [Google Scholar]
- Washenfelder, R.A.; Toon, G.C.; Blavier, J.F.; Yang, Z.; Allen, N.T.; Wennberg, P.O.; Daube, B.C. Carbon dioxide column abundances at the Wisconsin Tall Tower site. J. Geophys. Res 2006, 111, D22305, doi:10.1029/2006JD007154, 1–11.. [Google Scholar]
- Deutscher, N.M.; Griffith, D.W.T.; Bryant, G.W.; Wennberg, P.O.; Toon, G.C.; Washenfelder, R.A.; Park, S. Total column CO2 measurements at Darwin, Australia–site description and calibration against in situ aircraft profiles. Atmos. Meas. Tech 2010, 3, 947–958. [Google Scholar]
- Wunch, D.; Toon, G.C.; Wennberg, P.O.; Wofsy, S.C.; Stephens, R.S.; Fischer, M.K.; Zondlo, M.A. Calibration of the total carbon column observing network using aircraft profile data. Atmos. Meas. Tech 2010, 3, 1351–1362. [Google Scholar]
- Messerschmidt, J.; Geibel, M.C.; Blumenstock, T.; Chen, H.; Deutscher, N.M.; Engel, A.; Xueref-Remy, I. Calibration of TCCON column-averaged CO2: The first aircraft campaign over European TCCON sites. Atmos. Chem. Phys 2011, 11, 10765–10777. [Google Scholar]
- Andrews, A.E.; Boering, K.; Daube, B.C.; Wofsy, S.C.; Loewenstein, M.; Jost, H.; Strahan, S.E. Mean ages of stratospheric air derived from in situ observations of CO2, CH4, and N2O. J. Geophys. Res 2001, 106, 32295–32314. [Google Scholar]
- Messerschmidt, J.; Macatangay, R.; Notholt, J.; Petri, C.; Warneke, T.; Weinzierl, C. Side by side measurements of CO2 by ground-based Fourier transform spectrometry (FTS). Tellus B 2010, 62, 749–758. [Google Scholar]
- Sussmann, R.; Rettinger, M.; Borsdorff, T. The new TCCON-FTS site at Garmisch, Germany (47°N, 11°E, 744 m a.s.l.): Set up, first year of operation, and contribution to OCO and GOSAT validation. Proceedings of EGU General Assembly Conference, Vienna, Austria, 19–24 April 2009; p. 8780.
- Ohyama, H.; Morino, I.; Nagahama, T.; Machida, T.; Suto, H.; Oguma, H. Column-averaged volume mixing ratio of CO2 measured with ground-based Fourier transform spectrometer at Tsukuba. J. Geophys. Res 2009, 114, D18303, doi:10.1029/2008JD011465, 1–17.. [Google Scholar]
- Wang, D.Y.; Stiller, G.P.; von Clarmann, T.; Fischer, H.; López-Puertas, M. Cross-validation of MIPAS/ENVISAT and GPS-RO/CHAMP temperature profiles. J. Geophys. Res 2004, 109, D19311. [Google Scholar]
- Barkley, M.P.; Monks, P.S.; Engelen, R.J. Comparison of SCIAMACHY and AIRS CO2 measurements over North America during the summer and autumn of 2003. Geophys. Res. Lett 2006, 33, L20805. [Google Scholar]
- Wunch, D.; Wennberg, P.O.; Toon, G.C.; Connor, B.J.; Fisher, B.; Osterman, G.B. A method for evaluating bias in global measurements of CO2 total columns from space. Atmos. Chem. Phys. Discuss 2011, 11, 20899–20946. [Google Scholar]
- Reuter, M.; Buchwitz, M.; Schneising, O.; Heymann, J.; Bovensmann, H.; Burrows, J.P. A method for improved SCIAMACHY CO2 retrieval in the presence of optically thin clouds. Atmos. Meas. Tech 2010, 3, 209–232. [Google Scholar]
|Orbital altitude (km)||666||799.8||705.3|
|Spatial resolution (km)||10.5||30 × 60||13.5|
|Spectral region (μm)||0.75–14.3||0.24–2.38||3.7–15.4|
|Spectral resolution (cm−1)||0.2||1.0–7.8||0.55–2.0|
|Detecting instrument||FTS and CAI||8 channel grating spectrometer||grating spectrometer|
|Observation mode||nadir, flare, target||limb, nadir||nadir|
|Signal to noise ratio (dB)||120 (1.56∼1.72 μm)||<100 (1.57 μm)||2,000 (4.2 μm)|
|120 (1.92∼2.08 μm)||1,400 (3.7∼13.6 μm)|
|800 (13.6∼15.4 μm)|
|Site||Country||Coordinate [Latitude, Longitude]||Altitude [m a.s.l.]||Reference|
|Bialystok||Poland||53.23 °N, 23.02 °E||180|||
|Bremen||Germany||53.103 °N, 8.85 °E||30|
|Garmisch||Germany||47.47 °N, 11.06 °E||746.6|||
|Park Falls||USA||45.94 °N, 90.27 °W||442|||
|Lamont||USA||36.60 °N, 97.48 °W||320||[41, 44]|
|Tsukuba||Japan||36.05 °N, 140.12 °E||31|||
|Izaña||Spain||28.31 °N, 16.49 °W||2,373|
|Darwin||Australia||12.42 °S, 130.89 °E||32|||
|Wollongong||Australia||34.41 °S, 150.87 °E||30|
|Lauder||New Zealand||45.04 °S, 169.68 °E||370|
|Bremen (53.10°N, 8.85°E)||n||11||6||5||11||17||24||41|
|Garmisch (47.48°N, 11.06°E)||n||15||0||3||3||0||21||21|
|Park Falls (45.94°N, 90.27°W)||n||14||30||5||35||47||26||73|
|Lamont (36.60°N, 97.49°W)||n||26||5||11||16||6||30||36|
|Tsukuba (36.05°N, 140.12°E)||n||7||0||6||6||1||15||16|
|Izaña (28.31°N, 16.5°W)||n||20||0||0||0||14||24||38|
|Darwin (12.42°S, 130.89°E)||n||11||22||8||30||37||17||54|
|Wollongong (34.41°S, 150.88°E)||n||16||1||2||3||7||19||26|
|Lauder (45.04°S, 169.68°E)||n||4||28||3||31||54||18||72|
|Global Offset [ppm]||−6.53||−1.07||0.04||−0.78||0.21||2.81||1.84|
|Relative Precision [ppm]||1.77 (0.46%)||3.58||4.13||4.67 (1.22%)||2.19||2.56||2.76 (0.72%)|
|Relative Accuracy [ppm]||1.50 (0.38%)||2.51||3.19||3.35 (0.87%)||1.27||1.97||0.96 (0.25%)|
|Location||Annual Increase [ppm per year]||Seasonal Cycle Amplitude [ppm]|
|Global Offset [ppm]||−0.78||−0.70||1.84||1.81|
|Relative Precision [ppm]||4.67(1.22%)||4.12(1.07%)||2.76(0.72%)||2.68(0.70%)|
|Relative Accuracy [ppm]||3.35(0.87%)||1.65(0.43%)||0.96(0.25%)||1.01(0.26%)|
© 2013 by the authors; licensee MDPI, Basel, Switzerland This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).