2020-11-12

Regarding results of GOSAT-2 observational data analysis (carbon dioxide, methane, and carbon monoxide) and the start of their public release

(Distributed simultaneously to Tsukuba Science City Press Club, Environment Press Club, and Ministry of the Environment Press Club)

Thursday, November 12, 2020
Satellite Observation Center/GOSAT-2 Project,
National Institute for Environmental Studies, Japan
 MATSUNAGA Tsuneo Director
 YOSHIDA Yukio Senior Researcher
 MORINO Isamu Senior Researcher
 SAITO Makoto Senior Researcher
 NIWA Yosuke Senior Researcher
 OHYAMA Hirofumi Senior Researcher
 YASHIRO Hisashi Senior Researcher
 KAMEI Akihide Specialist
 KAWAZOE Fumie Specialist
 SAEKI Tazu Specialist
 

Greenhouse gasses Observing SATellite-2 “IBUKI-2” (GOSAT-2) was launched in 2018 and has been in nominal operation since February 2019. We have analyzed GOSAT-2 observational data using a full physics retrieval method, and retrieved global distributions of carbon dioxide, methane and carbon monoxide. This press release provides an overview of the retrieval results and their preliminary validation results. As for GOSAT-2 carbon dioxide retrieval results, this is the first report. The full physics retrieval results will be released to the public along with the proxy retrieval results after the relevant preparations, including documentation, are completed.
Note: The full physics retrieval method and proxy retrieval method are both methods for estimating greenhouse gas concentrations from satellite data but differ in the way influences of clouds and aerosols are treated. The former accounts for the influences more precisely using a physical model, whereas the latter processes them approximately with assumptions. For the retrieval of carbon dioxide, only the former method can be used.

Greenhouse gasses Observing SATellite-2 “IBUKI-2” (GOSAT-2) is a joint project promoted by the Ministry of the Environment of Japan (MOE), the Japan Aerospace Exploration Agency (JAXA) and the National Institute for Environmental Studies (NIES) (hereinafter, collectively, the “Three Parties”) as the successor to Greenhouse gasses Observing SATellite “IBUKI” (GOSAT), which was launched in 2009 and is still in operation.

GOSAT-2 was launched on October 29, 2018 from the JAXA Tanegashima Space Center with the H-IIA Launch Vehicle No. 40 (H-IIA F40). Later, GOSAT-2 collected the first light data with the onboard instruments: first on November 5-6, 2018, images with the Thermal And Near infrared Sensor for carbon Observation - Cloud and Aerosol Imager-2 (CAI-2), and on December 12-14, 2018, data with the Thermal And Near infrared Sensor for carbon Observation - Fourier Transform Spectrometer-2 (FTS-2), confirming that the onboard instruments were operating normally. Further, on February 1, 2019, GOSAT-2 shifted its operation mode to nominal and started global observations by CAI-2 and FTS-2. Note that the spectral radiance data observed by CAI-2 and FTS-2 have been distributed to the public as GOSAT-2 Level 1B Product from the following website since August 5, 2019:
GOSAT-2 Product Archive, National Institute for Environmental Studies
(https://prdct.gosat-2.nies.go.jp/en/index.html)

Subsequently, on July 5, 2019, NIES released the first retrieval results of methane (CH4) and carbon monoxide (CO) by a proxy-method (see http://www.nies.go.jp/whatsnew/20190705/20190705-e.html).

This press release provides an overview of the retrieval results of carbon dioxide (CO2), CH4 and CO by a full physics method. As for CO2 retrieval results, this is the first report from the GOSAT-2 observation. The full physics method is designed to calculate light propagation in the atmosphere based on a physical model, and used for GOSAT-2 to simultaneously retrieve column-averaged dry-air mole fractions of CO2, CH4 and CO. Among the greenhouse-gas-monitoring satellites currently in operation, GOSAT-2 is the only satellite capable of simultaneously estimating column-averaged concentrations of these three gas species.

Figure 1 shows the global distributions of column-averaged dry-air mole fractions of CO2, CH4 and CO (denoted by XCO2, XCH4 and XCO, respectively) retrieved from GOSAT-2 FTS-2 data in May 2019 using the full physics method.

Figure 1
Figure 1. Global distributions of XCO2 (upper left), XCH4 (upper right) and XCO (lower left) retrieved from GOSAT-2 FTS-2 data in May 2019 using the full physics method (in 2.5-degree mesh).

As seen from the whole-atmosphere monthly mean CO2 concentration based on GOSAT observation in May 2019 that hits the record high of 409.4 ppm at that time (see http://www.gosat.nies.go.jp/en/recent-global-co2.html), the photosynthesis of mid- and high-latitude forests in the Northern Hemisphere is not fully activated as of May 2019 and the CO2 concentrations in Russia and Canada remain high. In terms of CH4 concentration, it appears to be high in southeastern China, northern India, northern Europe/Africa and eastern North America while the concentration of CO, a gas species newly added for GOSAT-2 observation, seems high in the region from India through northern Southeast Asia to China, Central Africa, and Mexico and its leeward side. The high CO concentrations observed are considered to be attributable to the use of fossil fuels and biofuels (incl. charcoal) in the region from India to China, and to forest fires in Central Africa and Mexico.

Figure 2 shows the global distributions of column-averaged dry-air mole fractions of CO2 and CH4 retrieved from GOSAT FTS data in May 2019 using the full physics method. As seen from the figures, the areas, where GOSAT-2 data are available, have increased significantly compared to the areas with GOSAT data, including oceans and tropical regions with frequent cloud cover. The number of observed data from which CO2 and other gas species are successfully retrieved is 11,000 for GOSAT and 25,000 for GOSAT-2, which is 2.3 times the number of GOSAT. This increase owes mainly to the two technical improvements: the intelligent pointing mechanism —a new function introduced to FTS-2 that discriminates cloudy areas in real time using the camera installed inside the FTS-2 and automatically changes the observation point to a cloud-free area; and the extended range of possible pointing angle in the flight (along-track) direction in relation to sun glint observations over the ocean.

Figure 2
Figure 2. Global distributions of XCO2 (left) and XCH4 (right) retrieved from GOSAT FTS data in May 2019 using the full physics method (in 2.5-degree mesh).

The retrieval results of CO2 and other gases based on GOSAT-2 FTS-2 data using the full physics method have been validated with the ground-based observation data of the Total Carbon Column Observing Network (TCCON). By comparing FTS-2 data (land area) from March to May 2019 with ground-based data collected at 23 TCCON sites (Figure 3), a difference between these two data was found to be 4.1 ± 3.8 ppm (1.0 ± 0.9%) for CO2, 5.5 ± 17.4 ppb (0.3 ± 0.9%) for CH4, and 21.2 ± 9.0 ppb (22.7 ± 9.7%) for CO. (All these data from GOSAT-2 are higher than the TCCON data.) Although these values are better than GOSAT data two years after its launch, they have not reached the quality level of the latest data products from GOSAT FTS (V2.90, -0.35 ± 2.19 ppm (-0.1 ± 0.6%) for CO2, 2.2 ± 13.4 ppb (0.1 ± 0.7%) for CH4). Further improvement in the data products from GOSAT-2 is needed.

Figure 3
Figure 3. Comparisons of column-averaged dry-air mole fractions (vertical axis) of CO2 (upper left), CH4 (upper right), and CO (lower left) based on GOSAT-2 FTS-2 data by the full physics method with TCCON data (horizontal axis) (March-May 2019).

This result by the full physics method reported in this press release has been provided together with the result by the proxy method as GOSAT-2 Level 2 Product, to researchers whose research proposals have been adopted on the GOSAT Series Research Announcements (1st and 2nd RAs) and research organizations that have signed collaboration agreements with MOE, JAXA, and/or NIES. Once the related documents including validation results are ready, all these data products will be newly released to the general public from GOSAT-2 Product Archive.

We will work on further improvement in retrieval accuracy and increase in the number of retrieved data. Furthermore, we plan to proceed with inversion analysis of net fluxes, generation of Level 4 products, and estimation of anthropogenic emissions, and succeed to GOSAT’s results in a progressive manner.

[NIES staff in charge]

・Development of estimation algorithms for the column-averaged dry-air mole fractions of CO2, CH4 and CO using FTS-2 data: Yukio Yoshida, Senior Researcher
・Generation of products in operational data processing: Akihide Kamei, Specialist, Fumie Kawazoe, Specialist
・Validation using ground-based observations: Isamu Morino, Senior Researcher
・Generation of a priori data used for retrieving column-averaged dry-air mole fractions: Makoto Saito, Senior Researcher, Yosuke Niwa, Senior Researcher, and Tazu Saeki, Specialist
・Comparison with other satellite data in Europe and the US: Hirofumi Ohyama, Senior Researcher
・Analysis of CO concentration distribution: Hisashi Yashiro, Senior Researcher
・Project management: Tsuneo Matsunaga, Director for Satellite Observation Center

[Acknowledgements]

- For the generation of a priori data of CO2 and CH4 concentrations used for retrieving column-averaged dry-air mole fractions, a numerical simulation model, Nonhydrostatic ICosahedral Atmospheric Model (NICAM), was used to calculate atmospheric transport. For this calculation, sea-air CO2 flux data (1) and long-term reanalysis data, JRA-55 (2), provided by the Japan Meteorological Agency were used. The research and development related to these numerical simulations was supported by the Environment Research and Technology Development Fund (JPMEERF20172001) of the Environmental Restoration and Conservation Agency of Japan. A super computer system at the National Institute for Environmental Studies was used for the calculation (3).
 (1): https://www.data.jma.go.jp/gmd/kaiyou/english/co2_flux/co2_flux_data_en.html
 (2): https://jra.kishou.go.jp/JRA-55/index_en.html
 (3): https://www.cger.nies.go.jp/en/activities/supporting/supercomputer/

- The Research Computation Facility for GOSAT-2 (RCF2) of the Ministry of the Environment was used to develop a method for retrieving the column-averaged dry-air mole fractions of CO2, CH4 and CO reported in this press release.
 https://www.gosat-2.nies.go.jp/about/rcf2/

- The ground-based observation data described in this press release are obtained from a global observing network, the Total Carbon Column Observing Network (TCCON), which observes sunlight attenuated by atmospheric minor constituents when reaching the ground with ground-based Fourier transform spectrometers. The data in this study were obtained from the TCCON Data Archive hosted by CaltechDATA of the California Institute of Technology (https://tccondata.org).

[Contact]

Tsuneo Matsunaga
Director, Satellite Observation Center, National Institute for Environmental Studies

GOSAT-2 Project, National Institute for Environmental Studies
Phone: +81-29-850-2731/2966,
E-mail: gosat-2-info (please append ‘@nies.go.jp’ to complete the e-mail address)

[Glossary]

・ Satellite product:
Observation data by earth observing satellites are generally distributed by four major data products: Level 1 data products (equivalent to raw data), Level 2 data products (data converted into physical values), Level 3 data products (temporal and spatial averaged Level 2 data products), and Level 4 data products (processed data by inputting Level 1-3 data products into models). GOSAT-2 FTS-2 Level 2 data products are the column-averaged dry-air mole fractions of CO2, CH4, and CO. GOSAT-2 Level 4 data products are the net fluxes of CO2 and CH4 and the three-dimensional distributions of CO2 and CH4 concentrations.
・ Proxy method:
The proxy method is one of methods to reduce the effects of the variation of the optical path length caused by scattering by clouds and aerosols in the estimation of the column-averaged dry-air mole fractions. It uses a ratio of the column-averaged dry-air mole fractions of two gases whose absorption bands are adjacent to each other, which are retrieved under the assumption of no clouds or aerosols. It can obtain relatively accurate column-averaged dry-air mole fractions even if clouds and aerosols somewhat affect the optical path length, whereas it has disadvantages that its target gases of retrieval are limited, and CO2 is not applicable.
(Reference: Yoshida, Y. (2019) Journal of Remote Sensing Society of Japan, 39, 1, 22-28 (in Japanese))
・ Full physics method:
The full physics method is to retrieve the column-averaged dry-air mole fractions by directly calculating the effects of the variation of the optical path length using a forward model, with parameters which describes the status of clouds and aerosols. The full physics method has no limitation of its target gases of retrieval, whereas it is more easily affected by the variation of the optical path length caused by clouds and aerosols. Measures are necessary, such as the selection of data to be processed and the quality control of retrieved results.
(Reference: Yoshida, Y. (2019) Journal of Remote Sensing Society of Japan, 39, 1, 22-28 (in Japanese))
・ Column amount:
A numerical value of the total amount of gas represented by number of gas molecules in the column from the surface to the top of the atmosphere per unit area.
・ Column-averaged dry-air mole fraction:
A ratio of the column amount of a greenhouse gas to the column amount of dry air.
(Reference: https://www.nies.go.jp/kanko/kankyogi/69/column2.html (in Japanese))
・ Forward model:
The forward model described in this press release is a model to estimate the upward radiance spectrum at the top of the atmosphere, including sunlight reflected from the surface, observed by the satellite after the assumptions of gaseous concentration, cloud and aerosol distribution, and surface reflectance.
・ Optical path length:
The optical path length described in this press release is the product of the average length of sunlight propagating through atmosphere to the satellite multiplied by the refractive index of the atmosphere. Sunlight reaches the satellite through various routes while scattered by gaseous molecules, clouds and aerosols, and reflected from the surface. The routes vary depending on the amounts or distribution of clouds and aerosols and the surface reflectance. The optical path length also varies according to the route.
・ ppm (parts per million):
A kind of unit of atmospheric gas concentration. 1 ppm represents one-millionth.
・ ppb (parts per billion):
A kind of unit of atmospheric gas concentration. 1 ppb represents one-billionth.