Publications

Zhang, L; Ma, CF; Chen, X; Zhang, C; Li, Q; Ye, XM; Tian, LQ (2025). An integrated algorithm to estimate chlorophyll-a concentration in various optical waters using HY-3A CZI. ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 225, 402-422.

Abstract
The accurate retrieval of chlorophyll-a (Chla) concentrations in large-scale and highly-dynamic aquatic environments is challenging because of the complex and variable optical water properties, necessitating regionally-tuned or water-type-specific algorithms. In this study, we developed and validated an integrated algorithm to estimate Chla in optically diverse waters using the second-generation Coastal Zone Imager (CZI) onboard the newly launched HY-3A satellite (November 16, 2023). Based on a global in-situ dataset (N = 4,950, Chla range: 0.017-159.07 mg m(-3)), this algorithm first classified the dataset into three water groups using the remote sensing reflectance (R-rs lambda) ratio of 485 and 660 nm, and then applied optimal algorithms for each group: the ocean color index (OCI) algorithm for Group 1 (clear waters dominated by phytoplankton, where R(rs)485/R(rs)660 was >2.5); the normalized difference chlorophyll index (NDCI) and three-band index (TBI) blending algorithm for Group 3 (optically complex waters dominated by non-algal constituents, where R(rs)485/R(rs)660 was <1); and a weighted blending of the two for Group 2 (transitional waters, where R(rs)485/R(rs)660 was between these thresholds). By leveraging the strengths of these algorithms, model testing using in-situ data yielded high accuracy with an R-2 of 0.88 (in log scale), root mean square error (RMSE) of 12.79 mg m(-3), and average unbiased relative error (AURE) of 48.29 %, outperforming any single algorithm. Then, we applied the integrated algorithm to atmospherically corrected CZI R-rs for matchup validation and spatial mapping, further validating the algorithm performance and product quality. The independent validation with in-situ and satellite matchups over Lake Chaohu and Lake Donghu achieved a satisfactory accuracy, with an RMSE of 3.10 mg m(-3) and an AURE of 13.96 %. The CZI-derived Chla maps exhibited well-defined spatial continuity and heterogeneity over lakes and coastal waters, showing strong spatial pattern consistency with those from near-simultaneous Sentinel-2 Multi-Spectral Instrument (MSI) overpasses, further verifying the reliability of CZI-derived Chla products. As the first attempt to retrieve water quality parameters using HY-3A CZI, the integrated algorithm provides accurate and spatially continuous Chla retrievals across various optical waters, supporting the development of CZI water color products and their applications in aquatic environment monitoring and management.

DOI:
10.1016/j.isprsjprs.2025.05.001

ISSN:
1872-8235