Premium
An assessment of the accuracy and precision of water quality parameters retrieved with the Matrix Inversion Method.
Author(s) -
Campbell Glenn,
Phinn Stuart R.
Publication year - 2010
Publication title -
limnology and oceanography: methods
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.898
H-Index - 72
ISSN - 1541-5856
DOI - 10.4319/lom.2010.8.16
Subject(s) - colored dissolved organic matter , radiance , remote sensing , environmental science , inversion (geology) , water quality , dissolved organic carbon , overdetermined system , radiative transfer , hyperspectral imaging , soil science , mathematics , geology , chemistry , optics , phytoplankton , physics , environmental chemistry , paleontology , ecology , mathematical analysis , biology , organic chemistry , structural basin , nutrient
There is an increasing demand for quantitative inland water quality observations from satellites. One approach to estimate water quality constituent (chlorophyll and other pigments, total suspended material, and colored dissolved organic matter) concentrations from remote sensing radiance measured over inland waters is to apply a Matrix Inversion Method (MIM). Using the Hydrolight® radiative transfer model and typical water quality constituent concentrations from subtropical Wivenhoe Dam, Australia, this article demonstrates that significant improvements in the accuracy and precision of retrieved water quality constituent values can be obtained by using semianalytically estimated values for the proportionality factor that are calculated separately for each spectral band of the MERIS sensor. Furthermore, it shows that overdetermined systems of equations can be used to mitigate the effect of unknown and endemic sources of error in the remote sensing system. The residuals left after the inversion of the reflectance spectrum can also be used to assign a reliability measure to the retrievals of total suspended material (TSM) and colored dissolved organic matter (CDOM), but not for chlorophyll a. The results of this study may be used to improve algorithms for the remote sensing of water quality for freshwater impoundments.