11:15 AM - 11:30 AM
[ACG44-07] Exploring the optimal geometry of Himawari-8/9 to monitor vegetation dynamics across Southeast Asia
Keywords:Southeast asia, Himawari-8/9, Bidirectional Reflectance Distribution Function (BRDF), Observation Geometry, Vegetation Monitoring
Denese and widespread tropical rainforests across Southeast Asia play an important role in carbon cycling. At the same time, tropical rainforests are degrading rapidly due to population growth and exploitation of forest resources. Therefore, it is essential to understand vegetation dynamics across Southeast Asia. A new generation of Geostationary Earth Orbit (GEO) satellite, Himawari-8, and onboarding sensor, the Advanced Himawari Imager (AHI), has an observation frequency of 10 minutes and is expected to reveal vegetation dynamics across Southeast Asia with the high frequency of cloud cover.
In monitoring the Earth's surface using GEO satellites, it is extremely important to obtain optimal and uniform observation geometry over a wide area. The view geometry of GEO satellites is fixed, while solar geometry has diurnal and seasonal variations. Since reflectance can change depending on the positions of both the sun and sensor with respect to the surface, which is called the Bidirectional Reflectance Distribution Function (BRDF), applying the BRDF model to adjust the reflectance for certain geometry conditions is essential to remove the BRDF effects and capture actual variations in vegetation activity. However, the optimal geometry condition for Himawari-8 AHI across Southeast Asia has not been proposed.
This study compared four different geometry conditions and evaluated the reliability of each condition through correlative analysis between a two-band Enhanced Vegetation Index (EVI2) and site-measured Gross Primary Production (GPP) at six sites selected from AsiaFlux. The four geometry conditions include (i) Nadir condition, which is nadir view geometry and solar geometry at Local Solar Noon (LSN); (ii) LSN condition, which is fixed view geometry and solar geometry as LSN; (iii) Constant Scattering Angle (CSA) condition, which is fixed view geometry and solar geometry which meets the CSA criterion suggested by Gao et al. (2024); and (iv) Regional CSA condition, which is fixed view geometry and solar geometry which meets Scattering Angle of 140-150°. We found that the time-series of EVI2 under Nadir condition showed large uncertainty due to angular extrapolation, and that under LSN condition was strongly influenced by the hot spot conditions, where the solar geometry and the view geometry coincide. CSA condition remained BRDF effects depending on longitude. Finally, Regional CSA condition overcame these limitations of the other conditions, and EVI2 under Regional CSA had the highest correlation coefficients with GPP at almost all sites. The analysis results suggest Regional CSA condition to be the most desirable for vegetation dynamics studies with AHI in the tropical Southeast Asia region.
In monitoring the Earth's surface using GEO satellites, it is extremely important to obtain optimal and uniform observation geometry over a wide area. The view geometry of GEO satellites is fixed, while solar geometry has diurnal and seasonal variations. Since reflectance can change depending on the positions of both the sun and sensor with respect to the surface, which is called the Bidirectional Reflectance Distribution Function (BRDF), applying the BRDF model to adjust the reflectance for certain geometry conditions is essential to remove the BRDF effects and capture actual variations in vegetation activity. However, the optimal geometry condition for Himawari-8 AHI across Southeast Asia has not been proposed.
This study compared four different geometry conditions and evaluated the reliability of each condition through correlative analysis between a two-band Enhanced Vegetation Index (EVI2) and site-measured Gross Primary Production (GPP) at six sites selected from AsiaFlux. The four geometry conditions include (i) Nadir condition, which is nadir view geometry and solar geometry at Local Solar Noon (LSN); (ii) LSN condition, which is fixed view geometry and solar geometry as LSN; (iii) Constant Scattering Angle (CSA) condition, which is fixed view geometry and solar geometry which meets the CSA criterion suggested by Gao et al. (2024); and (iv) Regional CSA condition, which is fixed view geometry and solar geometry which meets Scattering Angle of 140-150°. We found that the time-series of EVI2 under Nadir condition showed large uncertainty due to angular extrapolation, and that under LSN condition was strongly influenced by the hot spot conditions, where the solar geometry and the view geometry coincide. CSA condition remained BRDF effects depending on longitude. Finally, Regional CSA condition overcame these limitations of the other conditions, and EVI2 under Regional CSA had the highest correlation coefficients with GPP at almost all sites. The analysis results suggest Regional CSA condition to be the most desirable for vegetation dynamics studies with AHI in the tropical Southeast Asia region.