9:30 AM - 9:45 AM
[ACG36-03] Development of Above Ground Biomass estimation method by combined use of GCOM-C / SGLI and Space LiDAR observation data
Keywords:GCOM-C/SGLI, GEDI, Above Ground Biomass
The SGLI sensor onboard the GCOM-C, launched in December 2017, provides observation data from nadir and slant observation in the Red and NIR regions. An algorithm has been proposed that uses a set of these observation data to estimate biomass using the observed reflectance change caused by the 3D structure of the vegetation canopy, and the biomass product created by this algorithm, currently available from JAXA.
In the current algorithm, the parameters of the biomass estimation formula are selected according to the forest type of the observed area. these parameters are adjusted by regression analysis using in-situ data from literature and existing biomass digital data.
Therefore, the abundance and accuracy of the biomass data used as the true value at that time are important. Currently, the data used as true values in regression analyzes are literature data collected in the past and existing digital biomass data. However, these data were collected and observed more than 10 years ago and are becoming less valid considering the growth of trees during that period.
On the other hand, biomass estimation technology from Space LiDAR observation data that can collect vegetation tree height information from space is developing. Currently, observation data is being collected by NASA's GEDI onboard the ISS from the end of 2019. In addition, JAXA is planning a new space LiDAR, MOLI which will be equipped with an optical imager. And it is considered that the biomass data estimated from these space LiDAR observation data can be used globally as the latest true biomass value. However, the LiDAR biomass estimation value has the following problems in incorporating it into the SGLI biomass estimation method.
It is necessary to verify the accuracy of the biomass estimate itself by LiDAR.A biomass product of NASA's GEDI is not available. JAXA is currently developing a biomass estimation algorithm for MOLI.The estimated biomass value by LiDAR is in the narrow area of footprint, and it is necessary to develop an algorithm that gives spatial representativeness in order to use it as the true biomass value in the observation area of a mediumresolution optical sensor.
In view of the above background, the purpose of this study is as follows.
The results of applying the biomass estimation algorithm for MOLI under development at JAXA to the fullwave data of the observation spot based on GEDI observation data will be verified by field measurement, and the estimation accuracy and validity will be examined.For GEDI data, generate global biomass estimates for observation spots using the above algorithm.Considering the data in 2) above as the true biomass value, perform SGLI biomass estimation and examine the superiority over the current algorithm. At that time, we will develop an algorithm for making the biomass of narrow area spots by LiDAR correspond to SGLI resolution.
Regarding 1), the validity and estimation accuracy of the algorithm were shown by conducting field observations on the already observed GEDI fullwave data and comparing it with the measured biomass. Regarding 2), typical land cover (using land cover classification data specialized for use in biomass estimation created using SGLI observation data), and the current biomass estimation algorithm by LiDAR for each region. Created a dataset to which. Regarding 3), the biomass estimation value by Space LiDAR created. Regarding 3), the land cover data created in 2) is created at SGLI's 250 m resolution. Since SGLI biomass products are created with a resolution of 1 km, there are actually 16 pixels of land cover data within one pixel of the biomass product. The results of calculating the weighted average for the biomass estimated from the GEDI data using the area ratios of these land covers within the 1 km resolution were used for the regression analysis of the SGLI algorithm to create the biomass estimation data from SGLI.
In the current algorithm, the parameters of the biomass estimation formula are selected according to the forest type of the observed area. these parameters are adjusted by regression analysis using in-situ data from literature and existing biomass digital data.
Therefore, the abundance and accuracy of the biomass data used as the true value at that time are important. Currently, the data used as true values in regression analyzes are literature data collected in the past and existing digital biomass data. However, these data were collected and observed more than 10 years ago and are becoming less valid considering the growth of trees during that period.
On the other hand, biomass estimation technology from Space LiDAR observation data that can collect vegetation tree height information from space is developing. Currently, observation data is being collected by NASA's GEDI onboard the ISS from the end of 2019. In addition, JAXA is planning a new space LiDAR, MOLI which will be equipped with an optical imager. And it is considered that the biomass data estimated from these space LiDAR observation data can be used globally as the latest true biomass value. However, the LiDAR biomass estimation value has the following problems in incorporating it into the SGLI biomass estimation method.
It is necessary to verify the accuracy of the biomass estimate itself by LiDAR.A biomass product of NASA's GEDI is not available. JAXA is currently developing a biomass estimation algorithm for MOLI.The estimated biomass value by LiDAR is in the narrow area of footprint, and it is necessary to develop an algorithm that gives spatial representativeness in order to use it as the true biomass value in the observation area of a mediumresolution optical sensor.
In view of the above background, the purpose of this study is as follows.
The results of applying the biomass estimation algorithm for MOLI under development at JAXA to the fullwave data of the observation spot based on GEDI observation data will be verified by field measurement, and the estimation accuracy and validity will be examined.For GEDI data, generate global biomass estimates for observation spots using the above algorithm.Considering the data in 2) above as the true biomass value, perform SGLI biomass estimation and examine the superiority over the current algorithm. At that time, we will develop an algorithm for making the biomass of narrow area spots by LiDAR correspond to SGLI resolution.
Regarding 1), the validity and estimation accuracy of the algorithm were shown by conducting field observations on the already observed GEDI fullwave data and comparing it with the measured biomass. Regarding 2), typical land cover (using land cover classification data specialized for use in biomass estimation created using SGLI observation data), and the current biomass estimation algorithm by LiDAR for each region. Created a dataset to which. Regarding 3), the biomass estimation value by Space LiDAR created. Regarding 3), the land cover data created in 2) is created at SGLI's 250 m resolution. Since SGLI biomass products are created with a resolution of 1 km, there are actually 16 pixels of land cover data within one pixel of the biomass product. The results of calculating the weighted average for the biomass estimated from the GEDI data using the area ratios of these land covers within the 1 km resolution were used for the regression analysis of the SGLI algorithm to create the biomass estimation data from SGLI.