Remote-sensing of benthic chlorophyll: should ground-truth data be expressed in units of area or mass?

摘要

During low tide, field spectrometric data (350-1050 nm) were acquired from intertidal mudflats in the upper reaches of Sydney Harbour, after which samples of sediment were taken using a small contact core. A total of 103 spectra/sediment pairs of samples were acquired. In the laboratory, amounts of chlorophyll in the contact cores were determined spectrophotometri-cally. The proportion of fine sediment (< 63 μm), a major factor influencing bulk-density of sediment, was measured. The relationship between the remotely sensed observations and the amounts of chlorophyll was investigated by comparing chlorophyll expressed as a mass per unit area (area-normalised values) and as a mass per unit mass (content or weight-normalised values) with: (ⅰ) reflectance between 350 and 1050 nm and (ⅱ) a ratio of reflectance at 562 and 647 nm (R562/R647). The reflectance at wavelengths > 400 nm were found to be more tightly correlated with area-normalised chlorophyll (maximal correlation = -0.51 at 666 nm) than with weight-normalised chlorophyll (maximal correlation = -0.41 at 664 nm). The relationship between R562/R647 and area-normalised chlorophyll was stronger (R~2 = 0.66) than for weight-normalised chlorophyll (R~2 = 0.47). The residuals from the regression of weight-normalised chlorophyll on R562/R647 were strongly correlated with the mass per unit mass of sediment that was < 63 μm. There was no such relationship for residuals from the regression of area-normalised chlorophyll on R562/R647. The results demonstrate conclusively that chlorophyll measured as mass per unit mass of sediment, is inappropriate for ground-truthing remotely sensed observations of chlorophyll in intertidal benthic sediments.