Speaker
Description
The next ESA's Earth Explorer will be BIOMASS, its launch is estimated in 2025 and its primary aim is to collect Synthetic Aperture Radar (SAR) acquisitions to improve the understanding of global forest status and temporal evolution [1]. BIOMASS is also the first P-band spaceborne SAR, with operational repeat-pass polarimetric-interferometric capabilities. This gives sensitivity to the full tree structure, as well as new opportunities to investigate sub-canopy topography and subsurface. Repeat cycle is optimized to be robust to temporal decorrelation and coverage is global except for Europe and North America due to international regulations. Ionospheric effects and disturbances are considered tolerable and mitigated through dedicated processing, which accounts also for coregistration and topography compensation [2]. Although, the main operational phase is interferometric dual-baseline (global coverage taking about seven months), there will be a first experimental tomographic phase, during which seven passes are collected over each location to enable 3D mapping (global coverage taking about fourteen months).
In this presentation we discuss the evolution of scientific forest products generation, which started from first prototypes [1] and lead during the last three years to the current operational implementation [3]. We also highlight insights and perspectives, some of which form the basis for further evolutions. The three main forest products generated by BIOMASS will be Forest Disturbance (FD), Forest height (FH) and Above Ground Biomass (AGB).
FD development is still in progress. On one side polarimetric analysis has been used to study the effects of environmental changes, which can compromise change detection methods based on the polarimetric covariance matrix [4]. On another side, investigation focused on detecting changes in a series of forest/non-forest classifications. Forest/non-forest classification is also needed for FH and AGB processing. Regarding FH, the effort was mainly devoted to investigating the use of tomographic reflectivity to improve classic polarimetric-interferometric model inversion [5]. The evolution of AGB estimation [6] has been focused on refining implementation at global scale, by defining regional and forest class partitioning to help in tailoring the algorithm to differing forest circumstances. A semi-empirical scattering model relating canopy signal to AGB is inverted to obtain model parameters where external reference is available, to subsequently compute AGB over an entire region. Attention is focused on identifying the most suitable source of reference data, additional work has been performed regarding helpful ancillary data and ground rejection techniques.
Testing on campaign acquisitions reprocessed to BIOMASS resolution are also presented to support the discussion.
References
[1] F. Banda et al., “The BIOMASS Level 2 Prototype Processor: Design and Experimental Results of Above-Ground Biomass Estimation”, Remote Sensing 2020
[2] F. Banda, et al., “BIOMASS Interferometric Calibration Processor Design”, IGARSS 2023
[3] M. Pinheiro et al., “The Biomass Processing Suite (BPS): an Overview of BIOMASS Operational Processor and Products”, POLINSAR 2023
[4] A. Alonso-González et al., “Polarimetric SAR Time Series Change Analysis Over Agricultural Areas”, IEEE TGRS 2020
[5] R. Guliaev et al. "Forest height estimation by means of TanDEM-X InSAR and waveform lidar data." IEEE JSTARS 2021
[6] M. Soja et al., “Design and Parameter Estimation Robustness of the Global Above-Ground Biomass Estimation Algorithm for ESA’s 7th Earth Explorer Mission BIOMASS”, IGARSS 2023
