Turbidity in the amazon floodplain assessed through a spatial regression model applied to fraction images derived from MODIS/Terra

The objective of this paper was to estimate turbidity in the Curuai floodplain during high water level. Spatial regression models were developed using fraction images derived from a Linear Spectral Mixture Model (LSMM) applied to a MODIS/Terra image and turbidity in-situ data. As the turbidity in-situ data showed spatial autocorrelation, they had been divided into four spatial regimes (clusters). Thus, a spatial regression model was developed for each spatial regime. Through the Akaike information criterion (AIC) it was verified which spatial regime showed the best fit in the spatial regression model. The results showed that the best fit was presented by the spatial regime 4 (r
²
=0.80, p≪0.05). The spatial regression model developed for the spatial regime 4 was then applied to all floodplain lakes. Spatial regression models show potential for turbidity studies in aquatic systems for considering spatial dependence between samples.