CHLOROPHYLL-a ESTIMATION IN PRODUCTIVE WATERS: THE EFFECT OF BACKSCATTERING ON ALGORITHM ACCURACY

摘要

Reflectance in several bands of the red-NIR domain isrnused in algorithms for chlorophyll a assessment inrnturbid, productive waters. We explored the effect ofrnparticle light backscattering on performance of thesernalgorithms. The range of chlorophyll a was from 6.6 torn17.5 mg m~(-3). We found that backscattering coefficientrndeclines exponentially towards longer wavelength andrnlikely affects the reliability of 3-band model in which arnwavelength independent backscattering coefficient inrnthe range from 660 through 750 nm was assumed. Ourrnresults indicate that the 2-band red-NIR model, in whichrnthe ratio of the reflectance centered on 708 nm (Sentinelrnband Oa11) to that centered on 665 nm (Sentinel bandrnOa8) is utilized, provides sufficient, and often the bestrnestimate of chlorophyll a concentration.