A sensitivity analysis of the 'Transmittance-Reflectance' method for measuring light absorption by aquatic particles

A thorough sensitivity analysis of the 'Transmittance-Reflectance' (T-R) method for measuringin vivo light absorption by aquatic particles retained on glass-fiber filters has been carried out. Particular attention has been devoted to the error contribution by the variability of the optical properties of the GF/F filters used. The overall error of the measurement of the filter-retained particle optical density, ODs, has been evaluated as similar to0.002 (1sigma), with 0.0015 error contribution by the filter variability. The corresponding error of the optical density of the particle suspension, ODsus, has been obtained by differentiation of the equation expressing the experimental correlation (OD(sus)versus ODs); the error increases with the optical density value, from sigma= 0.0015 (for ODsus = 0.05) to sigma= 0.027 at the upper limit of the optical density range tested (ODsus = 0.59). The validity range of the experimental (ODsus versus ODs) correlation has been shown to extend to ODs = 0.75. The importance of the accurate correction for light backscattering performed by the T-R method has been investigated by comparison with results given by the standard 'Transmittance' (T) method; it has been shown that the coarse correction for light scattering included in the T method may cause serious errors in the measured optical density spectra of mineral detritus. This evidence confirms the need to resort to the T-R method for the measurement of particle samples of detritus-rich coastal waters. The paper summarizes the latest developments of the T-R method and includes an Appendix with asynopsis of the routine now used by the authors.