Improved Conceptual Three-Band Model for Chlorophyll-a Retrieval in Inland Case-II Waters

Jing P. Xu,Fang Li,Bai Zhang,Kai S. Song,Zong M. Wang,Dian W. Liu,Guang X. Zhang

DOI: https://doi.org/10.1117/12.813039

2008-01-01

Abstract:In terms of the importance and challenge of accurate Chlorophyll-a (Chla) estimation in inland Case-II waters, many empirical or semi-empirical algorithms are established to extract information on Chla concentrations from remote sensing reflectance . However, the assumption that the optical parameter of Chla specific absorption coefficient a*ph(λ) remains constant usually restrains their estimating accuracy, including the conceptual three-band model [Rrs-1(λ1)-(λ2)]×Rrs(λ3) developed originally for estimating Chla amounts in terrestrial vegetation recently. Therefore, in this paper, an improved conceptual three-band model with the correction of Chla specific absorption coefficient [Rrs-1(λ1)-Rrs-1(λ2)]×Rrs(λ3×)×a*ph-1(λ1) was presented to estimate Chla contents for Shitoukoumen Reservoir, as a typical example of inland Case-II waters. According to the optical characteristics of waters studied, spectral regions included in the model were tuned to eliminate other interferences such as the variability of Chla fluorescence quantum yield, which resulted in the optimal positions for , and at 668nm, 678nm and 717nm respectively. Compared with the previous three-band model, the improve model gave out a much better estimating performance with a high coefficient determination (0.92) and a low root mean squared error (0.88μgl-1). Although the findings underline the rationale behind the improved model, an extensive database containing data in different water conditions and water types is required to test the accuracy of the model.