1 Introduction
Scientific research and observational data show that the global climate has undergone changes with warming as the main feature in the past hundred years. According to the IPCC report, over the past 100 years, the average temperature of the atmosphere above the global land has risen by 1.53 °C, and the global average temperature has risen by 0.87 °C, while the concentration of CO2 has risen from 280 to 419 ppm [1]. The CO2 concentration has a good correlation with changes in atmospheric average temperature. With the development of industrialization, human activities, especially the combustion of fossil fuels such as coal, oil and natural gas, emitting a large amount of CO2 into the atmosphere every year, significantly enhancing the greenhouse effect and being the major driving factor of global warming. In order to mitigate global warming, the international community has actively explored low-carbon and green development methods. The Paris Agreement, signed into force in 2016, proposes that countries should reach carbon neutrality as soon as possible, limit the global average temperature to 1.5 °C by the end of this century [2], and plan to conduct a global carbon stocktaking every five years from 2023. The world's major developed countries have proposed carbon neutrality targets and plan to achieve carbon neutrality around 2050. As the world's largest developing country, China officially proposed the "double carbon" goal of carbon peaking before 2030 and carbon neutrality before 2060 in September 2020. In order to accurately assess the progress of carbon neutrality goals, it is necessary to obtain an accurate total amount and distribution of carbon emissions and carbon absorption. Existing carbon emissions and carbon absorption monitoring tools rely mostly on ground-based inventory statistical method or ground-based carbon monitoring tools, and still lack the capability to monitor carbon sources, carbon sinks and discrete point sources at global and regional scales. Since satellite monitoring has the characteristics of large-scale, continuous and uniform standards, the application of space borne instruments to monitor the concentration of greenhouse gases in the atmosphere and the carbon sinks on the ground can effectively compensate the uneven distribution of ground monitoring stations, and obtain macro, continuous and dynamic greenhouse gas concentration and photosynthesis intensity monitoring data, combined with atmospheric transmission mode and ground monitoring data, which can realize the inversion of global carbon fluxes and further realize the verification of carbon emission inventories of various countries, so as to obtain more accurate carbon sources and sinks monitoring results. As early as 2002, ESA and the NASA developed Envisat and Aqua satellites respectively, enabling preliminary monitoring of greenhouse gas concentrations. In 2009 and 2014, the successful launch of the Japanese GOSAT satellite and the American OCO-2 satellite unveiled the prelude to the development of high-precision greenhouse gas monitoring satellites. China has also developed its own greenhouse gas monitoring satellite missions including Tansat and GF-5 etc. As the traditional player, The USA is developing a GEO carbon monitoring mission, namely GeoCarb, ESA is developing a three-satellite constellation of carbon stocktaking satellite, namely CO2M and Japan continue to develop its existing GOSAT series satellites.
From the perspective of development trend, there are multiple deficiency in the development of existing carbon monitoring satellites, including large differences in resolution and means, insufficient data sharing policy and international cooperation, lack of constellation observation between satellites, large differences in data inversion methods, non-uniform standards, etc. This paper recommends the development of China's carbon source carbon sink monitoring satellites system, through multi-satellite flying in constellation, and the strengthening of international cooperation in data inversion and assimilation, etc., to improve the monitoring accuracy of carbon sources and sinks, and better serve the goal of global carbon neutrality.
2 Carbon peaking and carbon neutrality requirement
Carbon peaking is defined as the total amount of CO2 emissions reach the peak and entering a platform period, and then gradually decreasing. Carbon neutrality means that the total amount of carbon emissions is equal to the amount of carbon absorption such as vegetation photosynthesis and carbon storage (CCUS), so as to achieve “Net zero carbon emissions”.
To evaluate carbon peaking and carbon neutrality, it is necessary to obtain the total amount and distribution of carbon emissions and carbon absorption. The traditional carbon emissions and carbon absorption method uses the inventory statistical method, using the emission or absorption coefficients of different industries and products given in the IPCC National Greenhouse Gas Inventory Guidelines, combined with the statistical data of fossil fuels, industrial production, agriculture, forestry and other statistics by each country every year to calculate a country's carbon emissions and carbon absorption data. The IPCC National Greenhouse Gas Inventory Guidelines are maintained and published by the IPCC, including the 1996 edition, 2006 edition, 2013 wetland supplement and 2019 edition (IPCC 2019).
2.1 Statistical method for carbon emissions monitoring
According to the IPCC National Greenhouse Gas Inventory Guidelines, the statistical categories of carbon emissions are roughly divided into 4 categories, as shown in Fig. 1.
Fig. 1 The category of greenhouse gas inventory |
In Fig. 1, the “Energy” means the emission of carbon by the burning up of the fossil fuels for energy production. The “industrial process and product use” stands for the emission of carbon by the product manufactory process and usage process. The “Agriculture, forestry and other land use” means the emission of CO2 and CH4 during the farming, forestry activities and other land use applications. The “waste” stands for the emission of waste disposal, burning the treatments.
The calculation formulas and coefficients for typical energy and industrial process carbon are shown in Table 1.
Table 1 The carbon emission calculation formula of energy and industrial process |
| Typical process | Formula | Coefficient | |
|---|---|---|---|
| Items | Reference Ci(kgCO2/kg) | ||
| Energy consumption | $ \mathrm{E}_{\text {energy }}=\sum_{i=1}^{n} Q_{i} \times C_{i}$. | Raw coal | 1.9003 |
| Coking coal | 2.8604 | ||
| Crude oil | 3.0202 | ||
| Nature gas | 2.1622 | ||
| Industrial process | $ \mathrm{E}_{\text {industry }}=\sum_{i=1}^{n} P_{i} \times C_{i}$ | Electricity | 1.2460 |
| Pig iron | 1.35 | ||
| Synthesis ammonia | 1.46 | ||
| Cement | 0.376 | ||
| Sodium carbonate | 0.138 | ||
Where Qi stands for the gross quantities of the specific kind of energy, Pi stands for the gross product of specific kind of industrial product.
The inventory statistic method can estimate a country's carbon emissions based on total energy consumption and production of industrial goods etc., but it is hard to consider the efficiency between different regions and get the distribution within a country. In order to achieve the control of carbon sources and sinks, it is necessary to monitor the emission intensity and distribution of each carbon emission source, and assess the photosynthetic intensity and the trend of the main carbon sink areas.
2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories for the first time proposed a complete method for the inversion of greenhouse gas emissions based on atmospheric remote sensing, which is a top-down method that can be used to verify the results of traditional bottom-up inventories statistics. By detecting the global CO2 gas concentration through satellite remote sensing, the carbon emission of each region can be retrieved from top to bottom, with the advantages of high precision, large-scale, continuous and dynamic, and the intensity and distribution of regional carbon dioxide and other greenhouse gas emission sources can be obtained, supporting the monitoring and supervision of carbon emission sources, as well as the joint prevention and control of greenhouse gases and polluting gases, providing support for the coordinated treatment of regional air pollution and greenhouse gases control. A number of greenhouse gases reversion algorithm have been developed for different missions (Shown in Sect. 3.1.1). Fluxes of CO2 cannot measure directly, atmospheric chemistry models are needed to inverse CO2 fluxes from CO2 concentration observed by satellite (Basu 2013; [3,4,5]. The accuracy of a CO2 emissions depends both on the concentration measurements and the “model inversion algorithm” to derive from the concentrations to the emissions. Assessment of anthropogenic emissions is difficult because the CO2 emitted by carbon sources is only a small fraction of the natural emission and absorption processes. A simplified estimation of the downstream concentration enhancement is given by Eq. 1 [6].
$\Delta =\frac{{F}_{emit}}{Area}\times {T}_{\mathrm{dwell}}\times \frac{1}{{\rho }_{column}}=0.4ppm\left(\frac{{F}_{emit}}{1\mathrm{Pg}/yr}\right)\times {\left(\frac{A}{{10}^{7}{\mathrm{km}}^{2}}\right)}^{-1}\times \left(\frac{{L}_{path}}{3\times {10}^{3}\mathrm{km}}\right)\times {\left(\frac{{v}_{wind}}{5m/\mathrm{s}}\right)}^{-1}\times \left(\frac{{\rho }_{column}}{3.6\times {10}^{6}{\mathrm{mol}/\mathrm{m}}^{2}}\right)$
where Δ is the concentration variation, typically given in ppm, Femit is the total emission from a region with area A. The dwell time of the measured air in the emitting region is Tdwell = Lpath/vwind, where Lpath is the path length and vwind is the average wind speed. The total atmospheric column density, ρcolumn is the column density appropriate to the total column depth of the atmosphere.
Remote sensing means can independently retrieve carbon emissions, obtaining third-party carbon emissions data, and can be used to check the greenhouse gas inventory statistics of countries and regions. With the continuous improvement of the greenhouse gas observation network and the stable operation of a number of high-precision greenhouse gas satellites in orbit, a large number of successful applications have emerged internationally. Researchers found that the results of CO2 concentration inversion in the central United States were in good agreement with the Greenhouse Gas Inventories [7]. Studies in the United Kingdom found that corrections were needed to revise the emission factors of IPCC air conditioners in cars [8]. The results of the inversion of CH4 emissions in Europe were higher than those of the inventory statistics, and there were errors caused by natural sources such as wetlands [9]. With the maturity of remote sensing retrieving methods, WMO has developed the IGI3IS (Global Greenhouse Gas Integrated Information) system, which evaluates carbon emissions, predicts development trends, and tests the effect of emission reduction based on remote sensing observations [10].
2.2 Statistical methods for carbon absorption monitoring
Carbon sinks mainly include carbon absorption in forestry, grassland, marine and agricultural production, as well as CCUS (Carbon Capture, Utilization and Storage) methods. The traditional carbon absorption statistics method adopts the inventory statistics method. One way to get the carbon sink can be approximately calculated from Eq. 2, with CCUS excluded [11].
$\mathrm{C}={C}_{f}{A}_{f}+\sum_{i}{C}_{i}{Y}_{i}/{\alpha }_{i}$
where C stands for carbon absorption rate, A stands for area, Y stands for yield of cropland, α stands for the economy coefficient rate of specific crop.
Another approach is to estimate carbon sinks in terrestrial and marine ecosystems using ecological process modelling methods. Taking the results estimated by the terrestrial ecosystem carbon sink model as an example, the estimation results of different dynamic vegetation models vary greatly, and the global terrestrial ecosystem carbon sink intensity simulated by 16 different models from 2007 to 2018 has a huge variation of 0.28-5.82 Pg C/a [12].The huge uncertainty in global carbon sources and sinks stems from both theoretical and cognitive flaws in carbon cycle models, as well as the lack of observational data with fine spatiotemporal resolution [4].
Since satellite remote sensing has global, large-scale and other characteristics, satellite remote sensing can be used to estimate biomass such as forests, grasslands, agriculture, seaweed, etc., and then estimate the amount of carbon absorption. Compared with the traditional ground-based carbon sink monitoring methods and inventory statistical method, satellites have the advantages of wide coverage, stability and continuity, and can provide more comprehensive and timely data for carbon emission management and emission reduction effect assessment, and have become the mainstream carbon sink assessment methods.
2.3 Carbon peaking and carbon neutrality overview
From the 1970s to the present, global carbon emissions have basically shown a correlation with global economic development, mainly because economic growth has increased the demand for electricity, oil and other energy sources in various economic sectors. At present, developed countries have completed industrialization, from high-energy heavy industrial production to high-end manufacturing, software and service industries with low energy levels, and have basically achieved carbon peaking. Developing countries are still generally in the process of industrialization, and the energy consumption level per GDP is much higher, and there is still a certain gap between carbon peaking.
As the largest developing country, China has taken the initiative to announce to the world plans for carbon peaking and carbon neutrality, showing the image of responsible. At present, two non-industrialized countries have declared carbon neutrality, 13 countries or regions have achieved carbon neutral legislation, 3 countries are in the process of legislation, and another 53 countries have issued policy declaration documents. Table 2 gives the carbon neutrality timetables published by the major countries.
Table 2 Time of Net Zero Emission Commitment |
| Country | Time of Net Zero Emission | Remark |
|---|---|---|
| Suriname | - | Achieve |
| Bhutan | - | Achieve |
| Germany | 2045 | In Law |
| Sweden | 2045 | In Law |
| European Union | 2050 | In Law |
| Japan | 2050 | In Law |
| United Kingdom | 2050 | In Law |
| France | 2050 | In Law |
| Canada | 2050 | In Law |
| South Korea | 2050 | In Law |
| Spain | 2050 | In Law |
| New Zealand | 2050 | In Law |
| Hungary | 2050 | In Law |
| Luxembourg | 2050 | In Law |
| Ireland | 2050 | Proposed Legislation |
| Chile | 2050 | Proposed Legislation |
| Fiji | 2050 | Proposed Legislation |
| Finland | 2035 | In Policy Document |
| Austria | 2040 | In Policy Document |
| US | 2050 | In Policy Document |
| Russia | 2060 | In Policy Document |
| China | 2060 | In Policy Document |
| India | 2070 | In Policy Document |
From the above analysis, it can be seen that the traditional inventory statistics method can obtain the total carbon emissions and carbon absorption of each country, but it is difficult to obtain carbon sources and sinks at regional level. Satellite remote sensing means can obtain the total amount and distribution of carbon sources and carbon sinks at the same time, and with the characteristics of fast, large-scale, unified standards, etc., in addition to monitoring the carbon source carbon sinks of human habitual residence, it can also monitor the sparsely populated areas such as the North and South Poles, oceans, deserts, etc., truly realizing high-precision, global coverage and efficient monitoring, which is an important means for future carbon sources and sinks monitoring and carbon stocktaking.
3 The development of Carbon remote sensing monitoring methods
Carbon remote sensing includes carbon source monitoring and carbon sink monitoring, which can be divided into two categories: passive detection and active detection according to the detecting source.
3.1 Carbon sources monitoring
3.1.1 Passive remote sensing
For carbon source monitoring, passive detection means belong to hyperspectral observation methods. There are two types of passive hyperspectral sounder for carbon detection distinguished by wavelength, thermal infrared and short-wave infrared. By thermal infrared spectrograph, satellites can measure the thermal emission of the atmosphere both day and night. Due to the disturb of water vapor and weighting functions, thermal infrared hyperspectral mostly detect the mid-tropospheric CO2 result rather than total column [6]. The other type of hyperspectral sounder applies the short-wave infrared spectrograph to detect the spectrum of sunlight reflected on the ground after being absorbed by greenhouse gases, and then retrieve the total column of greenhouse gases. Figure 2 shows the spectra of the O2-A channel, CO2-1 channel, CH4 channel and CO2-2 channel obtained by the GF-5 satellite Greenhouse gases Monitoring Instrument [13].
Fig. 2 The spectral of GF-5/GMI. a GMI Level 1 product of O2-A. b GMI Level 1 product of CO2-1. a GMI Level 1 product of CH4. b GMI Level 1 product of CO2-2 |
Europe, the United States and Japan apply satellite remote sensing scheme to monitor global carbon sources and carbon sinks, and build a LEO-orbit carbon monitoring system, and further develop high accuracy carbon monitoring systems and GEO-orbit carbon monitoring systems.
As early as 2002, NASA and ESA developed wide-band spectrometers to study the concentration of the atmosphere, including Envisat/Sciamachy, Aqua/AIRS, Aura/TES, and Metop/IASI. They are not specifically designed for greenhouse gases detection, but after the updating of the retrieval algorithm, they can all detect the precious result of greenhouse gases. But due to insufficient spectral resolution and the limitation of spectral band, it is difficult to meet the accuracy requirement in climate research. The CO2 total column measurement accuracy of Sciamachy is about 3 ~ 10 ppm [14,15,16]. The CO2 mid-troposphere measurement accuracy of AIRS is about 2 ppm, and about 2 ~ 20 ppm for IASI [17,18] and about 5 ppm for TES [19].
In order to meet the high accuracy total column greenhouse gases detection requirement, Japan and the United States have dedicated developed the greenhouse gas detection satellite GOSAT and OCO, respectively.
Japan’s GOSAT-1 was launched in 2009 and GOSAT-2 was launched in 2018. The satellite adopts the interferometer discrete dot detection method, the spatial resolution is 10 km × 10 km, and the CO2 detection accuracy is gradually improved from the early 4 ppm to 0.5 ppm, and the CH4 detection accuracy is improved from 34 to 5 ppb [20]. Japan plans to launch the GOSAT-3 satellite in 2023 and switch from traditional FFT to grating spectroscopy to obtain wide-swath, continuous coverage of greenhouse gas observation with a maximum observation swath of 1,000 km [21].
The U.S. OCO-1, OCO-2, and OCO-3 satellites were launched in 2009 (launch failure), 2014, and 2019, respectively. Among them, the spatial resolution of OCO-2 satellites is 1.29 km × 2.25 km, and the swath is 10.6 km, the spatial resolution of OCO-3 satellites is 1.6 km × 2.2 km, and the swath is 10 km. With the improvement of algorithm, the CO2 detection accuracy of OCO-2/OCO-3 is about 1 ppm [22].
In 2017, ESA successfully launched the Sentinel-5p satellite, carrying a wide-swath hyperspectral atmospheric sounder that can simultaneously detect atmospheric components such as O3, NO2, SO2 and CH4, with spatial resolution of 7 km × 7 km and a swath of 2600 km, making it the widest methane observation satellite in orbit [23].
ESA plans to launch three CO2M satellites, with the first satellite scheduled to launch in 2025 to meet the global carbon stocktaking demand in 2028. The spatial resolution of CO2M is 2 km × 2 km, in a swath of 250 km, the CO2 targeted detection accuracy of 0.7 ppm, and the CH4 targeted detection accuracy of 9 ppb [24]. The increase of the spatial resolution and accuracy enable the satellites to detect the plumes of power plants [25,26]. Towards spaceborne monitoring of localized CO2 emissions, an instrument concept is presented and showed the potential for subsequent CO2 flux [27].Besides, the potential for global detection of methane plumes from individual point sources with the new generation of spaceborne imaging spectrometers was evaluated [28].
In addition to LEO satellites, the United States plans to launch the GeoCarb satellite in 2022 to continuously detect greenhouse gases in real time from the geostationary orbit to meet the needs of spatial and temporal changes monitoring in atmospheric greenhouse gases emissions. GeoCarb has a spatial resolution of 5 to 10 km, an observation range of 5800 km, and the detection channel of CH4, CO2 and CO, of which the CO2 detection accuracy is up to 2.7 ppm [29].
Apart from the global high accuracy greenhouse gases monitoring satellite, there are several commercial greenhouse gases monitoring program proposed over the past 5 years, aiming for the regional monitoring of for carbon emission of point sources. The most famous commercial carbon monitoring satellite is GHGSat, developed by GHGSat Inc., A Canadian commercial company, with 3 GHGsats in orbit. GHGSat applies a special working mode by staring a high emission point target guided by Sentinel-5p/TROPOMI. By staring mode, GHGSat can obtain multiple integration time to get enough energy for the spatial resolution of 50 m [30]. Since GHGSat-D entered orbit in 2016, more than 5,000 observations have been made. The inversion accuracy of CH4 column concentration in a single observation is 8% ~ 25%, and the detection flux threshold can reach 1.0 ~ 3.0 t/h [31].
Figure 3 shows the artist’s rendition of deployed GOSAT-2 and Sentinel-5p.
Fig. 3 Artist’s rendition of deployed GOSAT-2 and OCO-2. a The deployed diagram of GOSAT-2 (Credit: http://copjapan.env.go.jp/cop/cop24/pavilion/05/). b The deployed diagram of OCO-2 (Credit: https://dev.spacenews.com/wp-content/uploads/2014/08) |
Figure 4 shows the artist’s rendition of deployed Envisat and Sentinel-5p.
Fig. 4 Artist’s rendition of deployed Envisat and Sentinel-5p. a The deployed diagram of Envisat (Credit: https:// www.universetoday.com/wp-content/uploads/2012/05/). b The deployed diagram of Sentinel-5p (Credit: https://www.esa.int/ESA_Multimedia/Images/2017/06/) |
China is also actively developing carbon source satellites, The Gaofen-5 satellite equipped with GMI (Greenhouse gases Monitoring Instrument) has achieved CO2 detection accuracy better than 4 ppm and CH4 detection accuracy better than 20 ppb in-orbit [32,33]. Tansat has achieved CO2 detection accuracy better than 2 ppm in orbit [34,35]. The satellite is targeted to realize global, all-day, high-precision and multi-scale carbon source detection. In addition, China is also developing a small satellites constellation system of high-resolution and high-efficiency carbon source monitoring. Each small carbon source satellite has a swath of 30 km and a spatial resolution of 300 m. The 12 small satellites are fly in formation in the morning and afternoon Sun-synchronous orbit (SSO), which can realize 3 ~ 4 consecutive observations of key areas a day, and have the ability to splicing 120-150 km swath for key areas above 30° latitude.
Figure 5 shows the artist’s rendition of deployed GF-5A and GF-5B.
Fig. 5 Artist’s rendition of deployed GF-5A/B. a The deployed diagram of GF-5A. b The deployed diagram of GF-5B |
Global seasonal map of CO2 and CH4 by GF-5 are shown in Fig. 6.
Fig. 6 Global seasonal map of CO2 and CH4 by GF-5/GMI. a Global Map of XCO2 in the Spring of 2019. b Global Map of XCO2 in the Autumn of 2019 |
3.1.2 Active remote sensing
Active detection mainly applies lidar to retrieve the column concentration information of greenhouse gases such as CO2 and CH4 by using the differential absorption of the actively emitted on and off laser pulse. Lidar does not rely on sunlight, so it can work 24/7 in-orbit. IPDA (Integrated Path Differential Absorption Lidar) is based on the idea of differential absorption. By comparing the echo signals of the two laser beams Pon and Poff, the differential optical thickness is calculated to obtain the column concentration of CO2 with high accuracy [36]. Figure 7 shows the working diagram of atmospheric concentration sounding lidar [37].
Fig. 7 The sounding diagram of atmospheric concentration sounding lidar |
The atmospheric environment monitoring satellite (DQ-1) was launched on April 16, 2022, which is the first satellite in the world to adopt active lidar means to achieve high-precision detection of CO2 concentration, with expected accuracy better than 1 ppm. High-precision greenhouse gas comprehensive detection satellite (DQ-2) will launch in 2023, and will be the first satellite in the world to achieve high-precision and high-resolution detection of carbon dioxide with jointly active and passive means, with a maximum detection accuracy of better than 1 ppm and a spatial resolution of 3 km × 3 km.
The artist’s rendition of deployed DQ-1 and DQ-2 are shown in Fig. 8.
Fig. 8 Artist’s rendition of deployed DQ-1 and DQ-2. a The deployed diagram of DQ-1. b The deployed diagram of DQ-2 |
The United States and Europe are actively developing the space borne active detection methods for high-precision detection of greenhouse gases. The United States proposed the ASCENDS mission, and Europe proposed the A-SCOPE, MERLIN and other greenhouse gas satellite programs with active detection capabilities, of which the MERLIN satellite is now in the process of developing, with envisage launch date to be 2025 [37].
Table 3 shows the main technical specifications of the major carbon monitoring satellites.
Table 3 Major specification of international carbon monitoring satellite |
| Satellite/ Payload | State/ Region | Launch | Method | GHG Targets | Accuracy | GSD | Swath | Characteristics |
|---|---|---|---|---|---|---|---|---|
| Envisat/ Sciamachy | EU | 2002 | Grating | CO2, CH4, etc | 3-10 ppm (CO2) | 30 km | 1000 km | Wide swath total column |
| Aqua/AIRS | US | 2002 | Grating | CO2, etc | 2 ppm | 13 km | 1600 km | Thermal emission, mid-troposphere |
| Aura/TES | US | 2004 | FFT | CO2, etc | 2 ~ 20 ppm | 50 km | 50 km | Thermal emission, mid-troposphere |
| Metop/IASI | EU | 2007 | FFT | CO2, etc | 5 ppm | 100 km | 2200 km | Thermal emission, mid-troposphere |
| GOSAT-1/2 | Japan | 2009/2018 | FFT | CO2, CH4, etc | 1 ~ 4 ppm 20 ppb | 10 km | 790 km (Discrete) | Steady in orbit, along with TIR band |
| OCO-2/3 | US | 2014/2019 | Grating | CO2 | 1 ppm | 2 km | 10.6 km | Push-broom sounding, |
| Tansat | China | 2016 | Grating | CO2 | 2 ppm | 1 × 2 km | 20 km | Push-broom sounding, along with SIF |
| GHGsat | Canada | 2016/2021 | F-P etalon | CH4 | 1%(20 ppb) | 50 m | 12 km | Staring mode for regional observation |
| Sentinel-5p | EU | 2017 | Grating | CH4 | 16 ppb | 7 km | 2600 km | Wide swath and daily coverage |
| FY-3H | China | 2023 | Grating | CO2, CH4 | 2 ppm 20 ppb | 3 km | 100 km | Detection of CO2, CH4, CO |
| GF-5 A/B | China | 2018/ 2021 | SHS FFT | CO2, CH4 | 2 ppm 20 ppb | 10 km | 800 km(Discrete) | No moving parts interferometer |
| DQ-1 | China | 2022 | Lidar | CO2 | 1 ppm | 50 km | - | The first CO2 Lidar |
| DQ-2 | China | 2023 | Lidar, grating | CO2, CH4 | 1 ppm 20 ppb | 3 km | 84 km | The first active and passive CO2 sounding satellite |
| GOSAT-3 | Japan | 2024 | Grating | CO2, CH4, NO2 | 1 ppm (CO2) | 1-3 km/10 km | 90 km/1000 km | Wide-swath GHG sounding |
| CO2M | EU | 2025 | Grating | CO2 | 0.7 ppm | 2 km | 250 km | Wide swath CO2 measurement |
| A-scope | A-scope | TBD | Lidar | CO2 | 1 ppm | - | - | Passing ground prototype test, space model to be approved |
| Ascends | US | TBD | Lidar | CO2 | 1 ppm | - | - | |
| Merlin | US | 2025 | Lidar | CH4 | 9 ppb | - | - | The first CH4 Lidar |
| GeoCarb | US | 2022 | Grating | CO2, CH4 | 2.7 ppm (CO2) | 5 ~ 10 km | 5800 km | The first GEO carbon monitoring |
3.1.3 Advances in Atmospheric GHG Remote Sensing Algorithms
The CO2 satellite inversion algorithm is mainly divided into two categories: empirical algorithm and physical algorithm. The empirical algorithm applies a large number of observation samples for training, collects CO2 training profile samples and their matching satellite radiation values, and then establishes atmospheric sample profiles through regression methods such as least squares, statistical control, artificial neural network methods, etc. This method of regression equation with satellite radiation values has high computational efficiency. The core problem of the empirical algorithm is how to establish a set of atmospheric profile samples that can represent different seasons and different locations, while considering the accuracy of the forward radiative transfer mode and the influence of clouds and aerosols. In addition, the deficiency of empirical algorithm is that it cannot provide the average kernel function and error estimation matrix like the optimization algorithm [38]. By using a nonlinear inference scheme based on neural network, Crevoisier [17] retrieved the CO2 distribution over the ocean in the global tropics using IASI clear sky data. For a spatial resolution of 5° × 5° with a monthly time scale, the retrieval accuracy was about 2 ppm (∼0.5%). The inversion results of empirical algorithms can be used as a priori profiles for physical algorithms.
Physical algorithms mainly include two basic algorithms: Differential Optical Absorption Spectroscopy (DOAS) algorithm and Optimal Estimation Method (OEM) algorithm. The WFM-DOAS algorithm based on the development of differential spectral absorption technology is mainly based on SCIAMACHY sensor observation data to retrieve the vertical column concentration of CO2/CH4 and other greenhouse gases. The original WFM-DOAS had low CO2 retrieval accuracy due to its inability to effectively correct the scattering effects [39]. Barkley improved the above algorithm with a priori profile containing the real atmospheric state parameters, and developed the Full Spectral Initiation (FSI)-WFM-DOAS algorithm, which considered three aerosol modes in combination, the calculation process is more complicated [40]. Later, Schneising made some improvements to the WFM-DOAS algorithm to reduce the errors caused by aerosol and cloud scattering effects, and applied the improved algorithm to invert the XCO2 concentration of SCIAMACHY 2003 ~ 2009 with long time series, resulting in a global annual mean XCO2 increase of 1.80 ± 0.13 ppm yr−1, which is in consistent with the CarbonTracker result (1.81 ± 0.09 ppm yr−1) [41]. Heymann developed the BESD (Bremen Optimal Estimation DOAS) algorithm to invert and compare the data of SCIAMACHY and GOSAT satellites. The comparison results demonstrate the good consistency between SCIAMACHY and GOSAT XCO2, that a mean difference for daily averages of 0.60 ± 1.56 ppm (mean difference ± standard deviation) for GOSAT-SCIAMACHY (linear correlation coefficient (r = 0.82), 0.34 ± 1.37 ppm (r = 0.86) for GOSAT-TCCON and 0.10 ± 1.79 ppm (r = 0.75) for SCIAMACHY- TCCON [14,15].
The optimal estimation method needs to determine the cost function first, and then use different optimization strategies to minimize the cost function. Both GOSAT and OCO currently use optimal estimation algorithms [42,43]. The NIES-FP inversion algorithm was developed by NIES. Y. Yoshida et al. applied the GOSAT data to invert the global distribution map of CO2 and CH4 using the optimal estimation algorithm. Compared with the TCCON data, the V01 version of the retrieval algorithm showed a larger negative impact. Deviation and standard deviation (8.85 and 4.75 ppm for XCO2 and 20.4 and 18.9 ppb for XCH4, respectively), the V02 version further improves the accuracy of the algorithm to -1.48 ppm by optimizing the solar irradiance database, aerosol parameters, O2 absorption cross-section, etc. and -5.9 ppb [44]. The ACOS (NASA Atmospheric CO2 Observations from Space) algorithm was developed by the NASA team and is currently the standard algorithm for OCO-2 data products [42].
Based on the OCO prototype algorithm, Boesch et al. of the University of Leicester developed the UoL-FP (University of Leicester Full Physics) algorithm. The average deviation between the GOSAT results retrieved using the 3G version and the global TCCON sites is -0.20 ppm, with a standard deviation of -0.20 ppm. is 2.26 ppm, and the correlation coefficient is 0.75 [45]. The RemoTeC (Remote Sensing of Greenhouse Gases for Carbon Cycle Modelling) inversion algorithm was jointly developed by the Netherlands Institute for Space Research (SRON, Netherlands Institute for Space Research) and Karlsruhe Institute of Technology (KIT, Karlsruhe Institute of Technology) [46], Compared to TCCON XCO2, RemoTeC XCO2 has a precision of 2.5 ppm and a bias of 0.36% averaged over all stations [47]. S. Basu applied the RemoTeC algorithm for the first time for CO2 global carbon fluxes [48].
Aerosol and surface albedo are important factors that determine atmospheric scattering, and are the most important parameters affecting the inversion of greenhouse gases. Certain algorithms are applied in the above physical inversion models to suppress errors, which are embodied in the following: The ACOS algorithm characterizes the aerosol and the surface applying the extinction efficiency, single scattering albedo and surface albedo of the four aerosol types. The UoL-FP algorithm of the University of Leicester applies the optical thickness profile and albedo, The RemoTec algorithm of the Netherlands Institute for Space Research uses aerosol content, height distribution and scale distribution. In addition, the algorithm developed by Liu Yi et al. for the TanSat satellite applies aerosol and surface pressure [49,50]. The CO2 algorithm for the GMI payload for the GF-5 satellite applies the optical thickness spectrum corresponding to the surface albedo and specific aerosol types by Shi Hailiang and Ye Hanhan et al. [13]. The main differences between different algorithms are the method of correcting atmospheric parameters and surface parameters in the inversion process. However, for the next generation of global carbon inventory monitoring satellite retrieval, the traditional correction algorithm needs further research.
3.2 Carbon sinks monitoring
The global ecological carbon sink system mainly includes two categories: land and ocean. Land vegetation parameters mainly include Clumping Index (CI), Leaf Area Index (LAI), Fraction of Absorbed Photosynthetically Active Radiation (FAPAR), Fraction of Vegetation Cover (FVC), Canopy Chlorophyll Content (CCC), Maximum Carboxylation Rate (Vcmax), Solar-induced Chlorophyll Fluorescence (SIF) etc. The vegetation coverage characteristics can generally be observed by passive multi-spectral means. The photosynthetic productivity of vegetation is mainly obtained by detecting SIF, inverting the efficiency of vegetation photosynthesis, and then obtaining Gross Primary Productivity (GPP) and Net Ecosystem Production (NEP). The structural characteristics of vegetation are usually obtained by means of lidar, and vegetation parameters can also be obtained by adding specific laser wavelengths, such as crown width, diameter at breast height, etc. [51,52].
Europe and the United States have developed a series of passive remote sensing means including multi-spectral and hyperspectral to detect global biomass and vegetation Net Primary Productivity (NPP). Typical satellites include LANDSAT series, TERRA/AQUA-Modis, Sentinel-2, Sentinel-3, etc. with multi-spectral methods, as well as the fluorescence observation project FLEX in the Copernicus program with hyperspectral methods [53]. China is still relying on multi-spectral vegetation and ocean color detection means, including Ziyuan series, GF-1, GF-2, GF-6, HY-1 etc. In terms of fluorescence detection, the CM-1 (Terrestrial Ecological Carbon Monitoring Satellite) carries a 20 km-swath payload dedicated to fluorescence detection, which is about to launch in 2022.
In active detection ways, the United States launched the active laser measurement satellites ICEsat-1 and ICEsat-2 in 2003 and 2018 respectively, and Europe officially established the P-band SAR satellite program Biomass in 2017. In terms of laser detection, ZY-3 and GF-7 have laser ranging capabilities, and the CM-1 has a 3-beam laser height detection capability. The multi-beam lidar detection capability is still too weak to meet the requirement of carbon sink monitoring. China also lacks the P-band SAR satellite for biomass detection.
Fig. 9 Global maps of SIF by Tansat. a Global Map of SIF for July 2017. b Global Map of SIF for December 2017 |
Ocean is the largest carbon pool on the earth. Atmospheric CO2 dissolve into the deep sea, become sediments or converts it into inert inorganic carbon through biological photosynthetic carbon fixation (biological pump) and seawater carbonate system (dissolution pump). The monitoring of marine carbon parameters mainly focuses on the parameters related to phytoplankton photosynthetic carbon sequestration, including chlorophyll concentration and primary productivity, as well as the concentration of sea surface organic carbon. For ocean carbon parameters, the concentration of chlorophyll a is mainly obtained by multi-spectral or hyperspectral methods. For example, the ocean water color observation satellites OLCI and GOCI can obtain information such as seawater Chla, Colored Dissolved Organic Matter (CDOM) and other information [55,56]. Water temperature data is available from various satellite sensors such as MODIS and TIRS [57,58]. At the same time, the laser method is used to obtain the information of the scattering layer of the underwater profile, so as to obtain the distribution information of particles, Particulate Organic Carbon, phytoplankton, and fish in the seawater (Zhao et al.2014, [59].
Table 4 shows the main technical specifications of the major carbon sink monitoring satellites.
Table 4 Major specification of typical carbon sink monitoring satellite |
| Satellite/Payload | State/ Organization | Launch | Method | Targets | GSD | Swath |
|---|---|---|---|---|---|---|
| Aqua/Modis | US | 2002 | Multi-spectral | CI, LAI, FVC, SST, TSM etc | 250/500/1000 m | 2300 km |
| FY-3/ Mersi | China | 2017/ 2023 | Multi-spectral | CI, LAI, FVC, SST etc | 13 km/3 km | 2300 km |
| LANDSAT/ OLI, TIRS | US | 2004 | Multi-spectral | SST, Chla, CDOM TSM, etc | 15/30/100 m | 185 km |
| GOCI | Korea | 2010 | Multi-spectral | POC, Chla, TSM etc | 500 m | 2700 km |
| Tansat | China | 2016 | Hyperspectral | SIF | 1 × 2 km | 20 km |
| OCO-2/3 | US | 2014/2019 | Hyperspectral | SIF | 2 km | 10.6 km |
| Sentinel-2/ MSI | EU | 2015 | Multi-spectral | LAI, FVC, Chla etc | 10/20/60 m | 290 km |
| Sentinel-3/OLCI, SLSTR | EU | 2016 | Multi-spectral | SST, CDOM, Chla etc | 300 m 1 km | 1270 km 1420 km |
| ICEsat-1/2 | US | 2003/2018 | Lidar | LAI, PFAR, Forest Height | - | - |
| CM-1 | China | 2023 | Lidar Hyperspectral | Height, SIF | 0.1 m height, 1 km | 20 km |
| Flex | EU | 2024 | Hyperspectral | SIF | 300 m | 150 km |
| GF-5 | China | 2018/ 2021 | Hyperspectral Multi-spectral | Chla, SST, LAI, FVC etc | 30 m 20/40 m | 60 km |
Carbon sink monitoring satellites can provide global coverage, complete parameters, and consistent key parameters products for the carbon cycle, providing important basic data for the dynamic and refined evaluation of carbon sources and sinks in terrestrial ecosystems. At the same time, combined with the monitoring data of atmospheric greenhouse gas concentration with high temporal and spatial resolution, high precision, and high time frequency from carbon source satellites, through the global carbon assimilation system, the flux contribution of anthropogenic and natural sources can be effectively distinguished and quantified, so as to achieve the peak of carbon emissions, scientific evaluation of the effectiveness of carbon neutrality actions and carbon emission accounting provide data support.
In the future, the development of global carbon assimilation system is more and more important. On the basis of improving the quality and quantity of observational data, it is also necessary to improve data assimilation methods and atmospheric chemistry models [60].
3.3 The development trend analysis
Through the study of the development trend of carbon monitoring satellite domestic and abroad, the trend of carbon remote sensing monitoring satellites mainly includes the following aspects:
1) Carbon source monitoring has developed from discrete point observation method or narrow swath observation measure to wide swath and continuous coverage measure, from low spatial resolution to high spatial resolution.
The early GOSAT, FY-3D, GF-5, etc. all adopted the discrete point observation method, OCO and Tansat adopted the narrow swath observation measure. At present, the greenhouse gases monitoring satellites under development all adopted a wide-swath and continuous coverage means, with the observation swath ranged from 100 to 1000 km.
2) The elements of carbon sink monitoring are becoming more and more complete, and the spatial and temporal resolution is getting higher and higher.
The key parameters of the ecosystem carbon cycle are detected from elements such as ecosystem classification and vegetation coverage characteristics, and new detections such as vegetation photosynthesis, vertical characteristics, and biomass are added, which further makes up for the lack of detection elements, and the temporal and spatial resolution is getting higher and higher.
3) From passive detection to active and passive in combination
At present, the in-orbit carbon monitoring satellites domestic and abroad all applying the passive hyperspectral detection methods, the detection accuracy is limited, and it is impossible to achieve all-day monitoring. The Developing lidar, P-SAR and other active detection methods, jointly detect with passive sounding measures, promisingly to achieve high precision, high resolution, large swath and all-day carbon source carbon sink detection.
4) From LEO-orbit observation to LEO-GEO orbit joint observation
At present, all of the carbon monitoring satellites are LEO-orbit remote sensing satellites, mostly in sun-synchronous orbit, which can achieve global data acquisition. The time resolution of the LEO satellites is usually low, and can only get the results of monthly or seasonal average coverage. In recent years, the United States has proposed the development of GEO carbon monitoring satellites. Through the combine observation of GEO and LEO constellations, global monitoring and near-real-time observation of key areas can be realized, which can meet the observation needs of high-precision, high spatial and temporal resolution of carbon sinks of carbon sources monitoring.
4 Gap analysis of carbon sources and sinks monitoring
1) The detection accuracy needs to be further improved
Spectral signals such as CO2, CH4, and SIF are rather weak, and easy to be interfered by H2O, Aerosol and cirrus cloud etc., make is difficult to retrieve. China's existing CO2 detection accuracy can only reach 1 ~ 4 ppm, CH4 detection accuracy can only reach 20 ppb, and there is still a big gap between the requirement of CO2 detection accuracy of 1 ppm and the CH4 detection accuracy of 10 ppb. The lidar satellite DQ-1 launched in April, 2022 can achieve the detection accuracy of 1 ppm, but due to the lack of observation swath, the coverage efficiency is still difficult to meet the user requirements, and it is urgent to make use lidar data to calibrate the wide-swath passive detection result through active and passive joint means to achieve wide-swath and high-precision detection. The spatial resolution of fluorescence monitoring of CM-1 satellites under development in China is only 20 km, the spectral resolution is not high enough, and the detection accuracy of SIF is insufficient. Due to the lack of multi-beam lidar and P-SAR detection methods, it is impossible to invert the height of global forests and tree diameter breast height, and it is urgent to develop new active detection methods such as multi-beam lidar and P-SAR. In order to support climate change research and environmental management, it is necessary to explore advanced detection methods and inversion algorithms to further improve the detection accuracy.
2) The spatial resolution of carbon source monitoring is not enough, and makes it difficult to locate carbon emission sources
In order to support the supervision of regional carbon sources and sinks, and the assessment of energy conservation and emission reduction effects, and the locating accuracy of carbon sources and sinks are better than 1 km. Constrained by the payload detection capability, the current GF-5 satellite in orbit adopts discrete dot sampling observation method, which is difficult to locate the carbon emission source. For the satellite under developing, the spatial resolution of carbon sources and sinks monitoring is only 3 to 50 km, which cannot achieve the high-precision locating of carbon emission sources, so as the spatial resolution of carbon monitoring needs to be improved.
3) The swath is narrow and the timeliness is insufficient
Due to the limitations of the LEO orbit and observation swath of the sounding payloads, it is difficult to achieve continuous and dynamic monitoring of regional area inside the countries. The existing carbon source monitoring satellites can only reach the revisit cycle of 1 month, and the carbon sink monitoring revisit cycle is about 1 week, which cannot meet the daily national coverage observation requirements of "double carbon".
4) Lack of key parameters of carbon cycle in long-term series and refined global spatial scale
The existing global carbon cycle key parameter products at home and abroad are not enough to support the accurate estimation of carbon sources and sinks, some global carbon cycle key parameters are lacking, and most of the existing carbon cycle key parameter products have not long enough time series and high spatial and temporal resolution. In addition, the key parameters of the carbon cycle from different data sources have insufficient consistency due to prominent spatiotemporal scale effects or differences in definitions and connotations, and need to be normalized and integrated.
5) Governments pay more attentions to the improvement of atmospheric greenhouse gases observation, other than to the improvement of atmospheric inversion assimilation models.
The “top-down” inversion algorithm not only needs to fully assimilate as much atmospheric observational data as possible, but also needs to make improvements in a priori flux uncertainty, chemical transport model accuracy, etc. to meet the need to validate global and regional carbon emissions with atmospheric observational data.
6) There are deficiencies in international cooperation
In addition, there are deficiencies in international cooperation in terms of validation, cross-calibration, retrieval algorithms and data policies and timelines. There is also insufficient participation in the formulation of international carbon monitoring standards.
5 Proposal for the development of carbon sources and sinks monitoring satellites
Due to insufficient spatial resolution and accuracy, China's carbon monitoring satellites both in-orbit and under development are difficult to locate the carbon emission sources, lacking the global carbon sources and sink monitoring capacity independent of inventory statistics. It is necessary to develop and launch a new generation of carbon monitoring satellite system, to achieve dynamic assessment of global greenhouse gases concentration and carbon sinks, supporting global carbon flux inversion, and the research of global climate change, enhancing China's contribution and influence in the field of carbon monitoring. It is recommended to make full use of the existing Ziyuan series resources satellites, Fengyun series meteorological satellites, Gaofen series satellites, as well as the Atmospheric Environment Monitoring Satellite etc., to improve the specifications of carbon sources and sinks monitoring, and building a multi-means, LEO-GEO constellation carbon monitoring satellite system to achieve higher precision, higher resolution and more dimensional carbon monitoring in the future.
It is proposed to carry out the developing of carbon monitoring systems in three stages, and achieve the capacity shown in Table 5 in 2025, 2030 and 2035, respectively.
Table 5 Development roadmap of carbon monitoring system |
| Milestone | Type | Targets |
|---|---|---|
| 2025 | Sources | Achieving 2 km resolution, 100 km swath, monthly global carbon source monitoring |
| Sinks | Achieving carbon sink monitoring by SIF monitoring, preparing for the technical readiness of P-SAR sounding | |
| 2030 | Sources | Achieving 2 km resolution, 500 km swath, global carbon source monitoring every 3 days |
| Sinks | Achieving wide swath SIF monitoring and P-SAR monitoring, capable of weekly monitoring of global carbon sinks | |
| 2035 | Sources | Developed geostationary carbon monitoring satellite, achieving daily carbon global monitoring and hourly area monitoring |
| Sinks | Achieving daily SIF monitoring, capable of P-SAR, multi-beam lidar for high accuracy carbon sink sounding |
Define the core satellites as the dedicated satellites for the carbon sources and sinks monitoring, and the auxiliary satellites as the integrated satellite with major mission in other applications. It is proposed to take to GEO carbon monitoring satellite and the new generation Environment Monitoring Satellite as the core carbon sources monitoring satellites, and to take the Fengyun series meteorological satellites and the hyperspectral series satellite as auxiliary. It is proposed to take the P-SAR satellite and the Global Biomass Monitoring Satellite as the core carbon sink monitoring satellite, and taking the Ziyuan series resources satellites and the Gaofen series satellites as auxiliary. The diagram of proposed carbon monitoring system is shown in Fig. 10.
Fig. 10 The diagram of proposed carbon monitoring system |
Major payload configurations and specifications of the core satellites are shown in Table 6.
Table 6 Major payload configuration and specifications of the core satellites |
| Satellites | Payloads configurations | Specifications | Coverage Capability |
|---|---|---|---|
| New generation Atmospheric Environmental satellites | CO2/CH4 Atmospheric Lidar Wide-swath GHG Spectrometer Cloud and Aerosol Imager Environmental Monitoring instrument | GHG: 2 km resolution, 100 km swath; Trace gases (NO2, SO2 etc.): 7 km resolution, 2600 km swath; Cloud and aerosol: 1.7 km resolution, 1900 km swath | GHG: monthly global coverage Trace gases: Daily coverage Cloud and aerosol: Near Daily Coverage |
| GEO Carbon Monitoring satellite | GEO GHG Monitoring Instrument GEO Environmental Monitoring instrument Cloud and Aerosol Imager | GHG: 2 km resolution, 1000 × 1000 km swath Trace gases (NO2, SO2 etc.): 5 km resolution, 3000 × 3000 km swath; Cloud and aerosol: 4 km resolution, 2000 × 2000 km swath | GHG: 5-h coverage, 1-h revisit; (East Asia) Trace gases: 2-h coverage Cloud and aerosol: 3-h coverage |
| Global Biomass satellite | Multi-beam Ranging Lidar Multispectral Imager Ultra-spectral SIF Spectrometer | Lidar: beam number up to 40 beams, ranging accuracy: better than 0.3 m; Imager: Band number better than 8 bands SIF: Spectral resolution better than 0.3 nm, spatial resolution better than 300 m | Monthly coverage |
| P-SAR satellite | P band SAR | Spatial resolution: 10 m × 5 m Swath: 190 km | Monthly coverage |
6 Conclusion
Based on the development status and trend study of carbon sources and sinks monitoring satellites and the demand of carbon peaking and carbon neutrality, it’s proposed to develop a multi-means, LEO-GEO constellation carbon monitoring satellite system in China. In order to achieve better coverage result and cross calibration opportunities, it is recommended that the LEO satellites fly in constellation with similar international satellites in orbit, carrying out retrieval algorithm research such as near-ground carbon dioxide gas concentration retrieval, global Net Primary Productivity (NPP) assessment, and global sea-gas carbon flux inversion etc. in advance, and make full use of the data from TCCON and other ground observation networks and the data of similar satellites in orbit to carry out validation tests and cross calibration, to improve the accuracy of data retrieval, and make full use of the satellites in carbon sources and sinks monitoring after the delivery of satellites in orbit. On this basis, improve the science data sharing policy internationally, actively participating in the formulation of international carbon monitoring standards, and improving the international influence of the carbon monitoring system. Through the application of different inversion algorithms and atmospheric assimilation models internationally, iteratively optimizes the satellite design, working mode and data processing algorithm to obtain the best application effect, establishing of an independent and controllable global carbon monitoring and assessment system, the monitoring and inventory verification of carbon emissions and carbon absorption in global regions at all levels, and contributing to the realization of carbon peaking and carbon neutrality.
Abbreviations
GMI:Greenhouse Gas Monitoring Instrument;GHGs:Greenhouse Gases;GF:Gaofen;FY:Fengyun;CHEOS:China High-Resolution Earth Observation System;AIOFM:Anhui Institute of Optical Fine Mechanics;CAS:Chinese Academy of Sciences;CCUS:Carbon Capture Utilization and Storage;TCCON:Total Column Carbon Observing Network;LEO:Low Earth Orbit;GEO:Geo-stationary Earth Orbit;NPP:Net Primary Productivity;IPCC:Intergovernmental Panel on Climate Change;P-SAR:P band Synthesis Aperture Radar;CI:Clumping Index;LAI:Leaf Area Index;FVC:Fraction of Vegetation Cover;CCC:Canopy Chlorophyll Content;Vcmax:Maximum Carboxylation Rate;SIF:Solar-induced Chlorophyll Fluorescence;NEP:Net Ecosystem Production;GPP:Gross Primary Productivity;NPP:Net Primary Productivity;CDOM:Colored Dissolved Organic Matter
Acknowledgements
The authors would like to thank all members of the Greenhouse Gas Monitoring Instrument (GMI) design and data processing team in Anhui Institute of Optical Fine Mechanics (AIOFM), Hefei, China, for their outstanding work in the development of GMI and the GHG retrieval systems, and to thank the China Center for Resources Satellite Data and Application (CRESDA), Beijing, China, for their guidance in the development of ground processing systems.
Authors’ Contributions
GM designed the constitution of the carbon sources and sinks monitoring satellite. YW analyzed the development status and trend of China’s carbon sources and sinks satellite and drafted the manuscript. MZ studies the algorithms of the GHG retrieval. YG participated in the carbon neutrality evaluation studies. XW provided the retrieval result of GMI in 2019. WJ studies the development status of international carbon monitoring satellite. NS provided the gap analysis of the carbon sources and sinks monitoring satellite. The author(s) read and approved the final manuscript.
Funding
The authors did not receive support from any organization for the submitted work.
Availability of data and materials
The data and material and available in the supplementary information.
Declarations
Ethics approval and consent to participate
The manuscript is original and not submitted to other journal, all the results are without fabrication. The authors consent to participate in the manuscript.
Consent for publication
The authors approve the version to be published.
Competing interests
The authors have no competing interests to declare that are relevant to the content of this article.
All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.
