Modeling of Physico-Chemical and Biological Parameters of Pao Cachinche Water Reservoir, Venezuela, using the Surface Reflectance from Landsat Satellite Images

Abstract

In this paper, it is proposed multivariable linear models to estimate the physico-chemical and
biological parameters of pao cachinche water reservoir using the surface reflectance from
landsat satellite images. Eight parameters are included: 1) Total Phosphorus, 2) Total
Nitrogen, 3) Plankton, 4) BOD, 5) COD, 6) Total Coliforms, 7) Electrical Conductivity and
8) pH. The results indicate that the adjustment between the water quality characteristics and
the surface reflectance extracted from Landsat satellite images are successful due to the RSquared statistic indicates that the models as fitted explain between 70.18 and 75.18% of the
variability in the physico-chemical and biological parameters. It has been found by each
model that only one spectral band might be removed because of the coefficient associated to
the recorded reflectances in this band has a low significant influence on the result of the
physico-chemical and biological parameters modeling.