Recently, Researcher He Xianqiang and collaborators from our institute published a research paper titled "An enhanced large-scale benthic reflectance retrieval model for the remote sensing of submerged ecosystems in optically shallow waters" in the top-tier journal ISPRS Journal of Photogrammetry and Remote Sensing (IF=12.7). The paper, with Dr. Wang Yuxin, a jointly supervised Ph.D. student with Zhejiang University, as the first author and Researcher He Xianqiang as the corresponding author, addresses the significant threat and degradation of shallow marine benthic habitats under the backdrop of climate change and intensified human activities.
The study focuses on the retrieval of bottom substrate spectral reflectance (Rb(λ)), crucial for indicating substrate type and health, amidst the challenge of satellite remote sensing signals (Rrs(λ)) coupling both water column scattering and seabed reflection without relying on prior knowledge. Building on their earlier work on the Semi-Analytical Benthic Reflectance retrieval model (SABR) (Wang, He* et al., RSE, 2022a), which integrated active lidar and passive high-resolution satellite imagery for benthic reflectance retrieval along satellite nadir tracks, this study introduces the Large-scale Benthic Reflectance retrieval model (LSBR).
Figure 1. Algorithm Flow of Large-Scale Benthic Spectral Reflectance Retrieval Model (LSBR).
LSBR advances beyond limitations of previous models dependent on lidar-derived water depth and water column attenuation coefficients, extending benthic reflectance retrieval from nadir tracks to large-scale areas solely based on remote sensing reflectance inputs. The LSBR algorithm leverages a comprehensive spectral library derived from field-measured benthic reflectance across diverse substrate types, identifying stable spectral relationships at 443nm and 490nm. Integrated with SABR and a previously developed large-scale water depth inversion algorithm (Wang, He* et al., 2022b), LSBR achieves remote sensing retrieval of both water chlorophyll concentration and benthic spectral reflectance in optical shallow waters.
Validation using Sentinel-2 satellite imagery in regions such as Xin Cun Port seagrass beds and Hua Guang Reef demonstrates LSBR's high accuracy compared to nadir track retrievals from joint active lidar-passive satellite images. Furthermore, analysis of temporal Sentinel-2 remote sensing data (2016-2023) over Canglu Reef illustrates LSBR's capability to detect substrate changes, effectively identifying areas of suspected degradation or recovery, showing good consistency with anomalies in sea surface temperature changes.
Figure 2. (A) Chlorophyll concentration obtained using (a) LSBR model and (b) Ocean Color Chlorophyll-a algorithm (OC3 algorithm) in Hua Guang Reef area; (B) Benthic reflectance distribution retrieved by LSBR model at (a) 443 nm, (b) 490 nm, (c) 560 nm, (d) 665 nm.
Figure 3. Analysis of typical regions of degradation and recovery in Canglu Reef.
The proposed LSBR model holds promise for optical shallow water benthic remote sensing monitoring in nearshore and offshore environments, enabling broad-scale application in substrate evolution analysis, substrate health monitoring, and carbon stock estimation. This research was supported by the Zhejiang Provincial Natural Science Foundation Major Project (Innovation Group) and other funding sources.
Citation: Wang, Y., He, X.*, Shanmugam, P., Bai, Y., Li, T., Wang, D., Zhu, Q., Gong, F., 2024. An enhanced large-scale benthic reflectance retrieval model for the remote sensing of submerged ecosystems in optically shallow waters. ISPRS Journal of Photogrammetry and Remote Sensing 210, 160–179. https://doi.org/10.1016/j.isprsjprs.2024.03.011
