Research and Development

2022.06.14 Tue

“SHIKISAI” (GCOM-C) marks the fifth year since its successful launch: Achievements in the terrestrial environment

1. About Global Change Observation Mission – Climate “SHIKISAI” (GCOM-C)
2. Detecting the essential information for photosynthetic activities estimation in terrestrial vegetation with high accuracy using GCOM-C
3. Monitoring land surface temperature at 250 m resolution
4. Challenges in estimating above-ground biomass (AGB) using the optical sensor
5. Acquiring in-situ ecological validation data on a large scale
6. Monitoring active volcanoes
7. Expectations for monitoring environmental changes by long-term observation

1. About Global Change Observation Mission – Climate “SHIKISAI” (GCOM-C)

The Global Change Observation Mission – Climate (GCOM-C), a Japanese nickname is “SHIKISAI,” was launched on an H-IIA rocket from JAXA Tanegashima Space Center in December 2017. “SHIKISAI” means “colorful” in Japanese. As the name implies, it captures the distribution and changes of vegetation, clouds, and “snow and ice” around the world in rich-vibrant colors. The name “SHIKISAI” conjures an image of the visible-light observation; however, the Second generation GLobal Imager (SGLI) onboard GCOM-C observes 19 multi-wavelengths of light, including near-ultraviolet to thermal infrared light. SGLI also observes the multi-directional light. Using such GCOM-C observation data, we can obtain valuable information regarding radiation, heat budget, and the carbon cycle.
This article, one of the series articles (see the articles here), introduces the GCOM-C’s fifth-year achievements and prospects in the terrestrial environment. GCOM-C land products are mainly followings: terrestrial ecological products (vegetation index (VI), above-ground biomass (AGB), leaf area index (LAI), etc.) based on multi-wavelength and multi-directional observations, and surface temperature products based on thermal infrared observations with a high spatial resolution (250 m resolution). GCOM-C terrestrial research group is working on monitoring the current state of the terrestrial environment with high accuracy, hoping to clarify the climate change mechanisms using GCOM-C data. The following sections introduce our efforts regarding terrestrial ecological products, active volcano monitoring, etc. Now that four-year observation data have been obtained from GCOM-C, the basis for clarifying the climate change mechanisms is finally in place.

> Link to the article about land, ocean, and snow and ice observations <

2. Detecting the essential information for photosynthetic activities estimation in terrestrial vegetation with high accuracy using GCOM-C

Global warming occurs when atmospheric concentrations of greenhouse gases, mainly carbon dioxide, increase. Since plants store carbon while absorbing carbon dioxide and generating oxygen during photosynthesis, it is crucially important to capture “the photosynthetic activity” on a global scale. GCOM-C observes nearly the entire globe in just two days and detects the essential information to estimate global photosynthesis in a timely way.
In order to estimate the amount of photosynthesis, it is important to estimate how much light, Photosynthetically Active Radiation (PAR), plants absorb. Using the Forest Light Environmental Simulator (FLiES), GCOM-C estimates the Fraction of Absorbed Photosynthetically Active Radiation (FAPAR) of terrestrial vegetation and the Leaf Area Index (LAI, green leaf area per unit ground surface area), related to photosynthetic capacity (Figure 1). The feature of the GCOM-C LAI product is to estimate the LAI in two layers (canopy and understory vegetation). Existing satellite sensor products (e.g., MODIS) estimate LAI in one layer (that is, all terrestrial vegetation from the ground surface). However, canopy and understory vegetations differ in seasonal changes in photosynthesis, etc., so understanding each LAI can help improve the accuracy of terrestrial ecosystem models and contribute to ecosystem research.
To estimate LAI with high accuracy, we continue to improve the model (e.g., improving the virtual forest used in FLiES). FLiESvox, the improved model of FLiES, can estimate the seasonal changes in LAI changes of the vegetation in the tundra area (in the red frame in Figure 2) and the coniferous forest in the subarctic region (in the yellow frame in Fig. 2), which are mostly distributed in the high latitudes of the Northern Hemisphere (Fig. 2).

Figure 1. GCOM-C products (data for ten days from July 11, 2021): (a) LAI of the canopy, (b) Total LAI including understory vegetation, (c) FAPAR. Figure by Dr. Toshiyuki Kobayashi (University of Tsukuba).
Figure 2: Estimation of LAI in the high latitudes of the Northern Hemisphere using FLiESvox. In the figures, the horizontal axis corresponds to longitude, and the vertical axis corresponds to latitude (map: sinusoidal projection). The upper image shows an example of LAI at latitude 50° to 90°N and longitude 0° to 360°. The lower images show the seasonal changes of LAI in Alaska, USA. Figure by Dr. Hideki Kobayashi, Japan Agency for Marine-Earth Science and Technology (JAMSTEC)

In order to estimate the amount of photosynthesis, it is also important to capture the period of photosynthetic activity. GCOM-C detects global plant phenology (i.e., the dates of leaf expansion, leaf fall, etc.) with high accuracy (Figure 3). In particular, GCOM-C detects the leaf-fall date with the highest accuracy among moderate-resolution satellite sensors (Figure 4). The plant phenology must change depending on extreme weather events and global warming. Therefore, it is necessary to continue the monitoring in the long term.

Figure 3: Global distribution of the date of leaves’ (a) Greenup onset, (b) maturity onset, (c) senescence onset, and (d) dormancy captured by GCOM-C. Figure by Dr. Wei Yang (Chiba University).
Figure 4: Comparison of the accuracy in the detection of plant phenology among three satellite sensors: the uppers are the date of leaf expansion, and the lowers are the date of leaf fall. The accuracy was validated with ground-based observation such as “Phenological Eyes Network” (PEN; You can find the best accuracy in the GCOM-C (see the highest coefficient of determination R2 and the smallest RMSE and bias). Figure by Dr. Wei Yang (Chiba University).

3. Monitoring land surface temperature at 250 m resolution

One of the GCOM-C features is to obtain land surface temperature (LST) with high resolution (250 m resolution), which enables us to detect even local phenomena such as the thermal belt on the slope of Mt. Tsukuba in Japan (Figure 5): In the daytime (upper left in Figure 5), the land surface temperature is lowest at the top of the mountain and higher at the foot depending on the elevation. In contrast, at night (the upper right figure), the temperature is higher at the mountainside than at the top and even at the foot of the mountain. Such a high-temperature area is called the thermal belt, caused by the specific physiographic feature that Mt. Tsukuba is an independent peak facing the Kanto Plain.

Figure 5: Thermal belt on the slope of Mt. Tsukuba detected by GCOM-C. Figure by Dr. Kenlo Nasahara (University of Tsukuba).

The GCOM-C LST depending on soil moisture content and plant water stress, with a high spatiotemporal resolution, can contribute to the effective and sustainable management of water resources, which might be threatened by future population growth and climate change. For example, at the Urmia Lake Basin in Iran, which faces existential difficulty, estimating evapotranspiration (ET) is crucial because water resource in a closed basin is determined by the balance of precipitation and ET (Figure 6). In Figure 6, ET was estimated using the GCOM-C LST and its behavior: LST is cooler in the water area and warmer in the non-water area (dry ground).

Figure 6: Evapotranspiration (ET) map at the Urmia Lake Basin in Iran estimated using GCOM-C land surface temperature in 2019. Figure by Dr. Masahiro Tasumi (Miyazaki University) and Dr. Masao Moriyama (Nagasaki University).

GCOM-C LST with a high spatiotemporal resolution enables us to estimate daily ET in each farm parcel (except for a smaller parcel than GCOM-C resolution) because of the increased frequency of usable clear day’s data. For example, the time series of ET estimated from GCOM-C LST enables us to monitor the irrigation condition without introducing expensive observation equipment in the field in each parcel of farm field in the western USA (Figure 7): High ET implies that irrigation is working effectively.

Figure 7: ET monitoring in the western USA, where large farms with the irrigation system are widespread. Figure by Dr. Masahiro Tasumi (Miyazaki University) and Dr. Masao Moriyama (Nagasaki University).

4. Challenges in estimating above-ground biomass (AGB) using the optical sensor

For global warming studies, it is important to estimate the amount of carbon storage in vegetation by photosynthesis (i.e., biomass) from satellites. Especially, trees play a great role in the storage of carbon in the long term because trees have a large volume and a long lifespan. Accordingly, estimating the distribution of above-ground biomass (AGB) and its amount is highly important for carbon cycle studies. Also, it is important to detect the variation of AGB because the emission of carbon stored in trees into the atmosphere, caused by deforestation, forest fires, etc., significantly impacts the carbon cycle.
The data of active sensors (such as satellite lidar and synthetic aperture radar [SAR]), which emit beams and detect the reflection from the earth’s surface, are generally used for the large-scale AGB estimation. Specifically, AGB is estimated from canopy heights acquired from the time lag between the reflected signals from the top of the tree and the ground surface.
GCOM-C onboard an optical (passive) sensor, which observes natural light reflected or emitted from the ground surface. A passive sensor is considered unsuitable for estimating accurate AGB because it has difficulty detecting tree heights (or canopy heights), which is important for AGB estimation. However, GCOM-C multi-directional observation made it possible to estimate accurate AGB by obtaining the vegetation-layer thickness, vegetation density, and vegetation coverage.
A problem with GCOM-C/SGLI AGB product version 2 (old algorithm) was the underestimation in the tropical region where large AGB exists. This fact was attributable to the rarely available in-situ data in this region. To overcome this problem, we applied the data of Global Ecosystem Dynamics Investigation (GEDI, lidar on board the International Space Station [ISS] by NASA) to the AGB estimation algorithm for the Multi-footprint Observation LIDAR and Imager (MOLI, planning to be installed on ISS by JAXA). Thus, we acquired much AGB data (instead of in-situ data). By adding those data, we achieved a high AGB estimation in the new algorithm for the GCOM-C/SGLI product (version 3) (Figure 8). The AGB estimation using satellite lidar and SAR tends to be saturated at over 150 t/ha. In contrast, GCOM-C AGB is not saturated up to 500 t/ha. Global AGB estimation with high accuracy using GCOM-C is expected to contribute to the carbon cycle research greatly.

Figure 8: Accuracy improvement in GCOM-C AGB estimation by adding the data acquired from GEDI data and MOLI algorithms. Figure by Dr. Yoshiaki Honda (Chiba University).

5. Acquiring in-situ ecological validation data on a large scale

JAXA is also working on acquiring reliable in-situ ecological data for validation. Terrestrial vegetation is a complicated system in tree species composition, canopy structure, and vegetation layers (i.e., tree layer, midstory layer, herb layer, etc.). Furthermore, its distribution is highly heterogeneous. Therefore, it is difficult to obtain reliable LAI at a large scale (such as a representative LAI corresponding to a GCOM-C pixel area (250 m × 250 m)). Furthermore, there are the following issues in in-situ data for validation: the discrepancy in an area between satellite and the in-situ observation and the geolocational discrepancy between satellite and in-situ observations. To overcome such difficulties and issues, A large-scale ecological observation project, “JAXA Super Sites 500,” was initiated. We evaluated and chose the most suitable observation method for a large-scale area (500 m × 500 m) and developed the uncertainty quantification method in the ecological validation data (Figure 9). The ecological validation data acquired at three forest sites and four grassland sites in this project were used for GCOM-C validation.

Figure 9: (Left) Location of observation sites of “JAXA Super Sites 500” acquiring reliable in-situ ecological validation data at a large scale (500 m × 500 m), (Right) Basic sampling design at a site. Figure modified from Akitsu and Nasahara (2022).

6. Monitoring active volcanoes

The GCOM-C terrestrial data is not only used for ecosystem monitoring but also used for monitoring active volcanoes in East Asia ( As shown in Figure 10, small-scale eruptions of Mt. Sakurajima were detected by GCOM-C, taking advantage of its features of high spatiotemporal resolution.

Figure 10. Small-scale eruption signals of Mt. Sakurajima, detected by GCOM-C in 2019. Left graphs show the detection of eruption signals by GCOM-C: The small-scale eruption signals could be detected by the change in brightness temperature at shortwave infrared (1.63 µm wavelength, upper image), although such activities could not be detected only by the change in brightness temperature at thermal infrared (10.8 µm wavelength, lower image). Figure by Dr. Takayuki Kaneko (University of Tokyo).

7. Expectations for long-term observation to monitor environmental changes

We are developing and providing reliable GCOM-C land products through the activities to acquire reliable validation data and the leading-edge algorithm development, as described above. GCOM-C data is useful for the real-time monitoring of water use in irrigation facilities and volcanic activity. Besides, GCOM-C is expected to detect global environmental changes caused by extreme weather events and global warming by providing the essential ecological information (such as LAI, AGB, and plant phenology) for the long term.
As for the total amount of carbon compounds produced by photosynthesis of plants in an ecosystem (gross primary production [GPP]), we are now developing an algorithm to estimate the value using the GCOM-C’s ecological information and land surface temperature (Figure 11), the figure shows that GPP in summer was higher in 2019 than the mean in 2011–2015 at high latitudes. This fact may detect the global warming event. Unfortunately, this event was derived from comparing the estimations using GCOM-C data in 2019 and the other sensor’s data in 2011–2015. It is not enough to detect such long-term global warming events only using GCOM-C four years’ data. We believe that GCOM-C long-term data can largely contribute to clarifying the phenomena caused by global warming in the future.

Figure 11. Estimation of gross primary production (GPP) in Asia: (Left) GPP on DOY209 in 2019 obtained from GCOM-C data (DOY means the day of the year. In 2019, DOY209 is July 28), (Right) The mean GPP on DOY209 in 2011–2015 obtained from MODIS data. Figure by Dr. Kazuhito Ichii (Chiba University).

Writings: JAXA/EORC, Dr. Tomoko K. Akitsu

Stay Connected