Spatio-temporal variation of nitrate based on Landsat 8 in Playa Colorada bay, Sinaloa, Mexico.

Abstract

Characterizing water quality in coastal waters through spatial observations is challenging, due to spatial and temporal variations in water composition. Nitrate, an important compound for water quality assessment, has received little attention in estimates made from satellite measurements, even though it can be estimated using models generated from multispectral images. Since nitrate is a non-optically active parameter that can be correlated with optically active parameters, it was related to bands of the visible and infrared spectrum, captured in Landsat-8 images, and used to generate empirical models to estimate the spatio-temporal variation of nitrate concentration in the Playa Colorada Bay, in the state of Sinaloa Northwest Mexico. Four sampling campaigns were performed, two in spring and two in fall. Nitrate nitrogen (NO3-N) concentration ranged between 0.69 and 1.80mg/L, values higher than those recommended in the Mexican ecological criteria of water quality for the protection of marine aquatic life in coastal areas. Generated models showed a significant relationship (P < 0.05) between NO3-N and band reflectance in the infrared (band 5) and short-wave infrared (band 6 and band 7) spectra of Landsat-8 imagery. The B6 band appeared in all models selected to estimate NO3-N in the bay. These results evidence the potential of Landsat-8 images for the estimation of nitrate in the coastal waters of Sinaloa, México.