Water quality assessment in a wetland complex using Sentinel 2 satellite images

Abstract

This research was carried out in a wetlands complex of great environmental and socioeconomic importance located in the north of Colombia, highly impacted by different anthropic activities that take place in the region, e.g., livestock, agriculture and urban wastewater disposal. Field measurements of pH, temperature, turbidity, electrical conductivity and dissolved oxygen were performed at 22 sites located in the Bajo Sinu wetlands complex. This wetlands complex includes Cienaga Guatinaja, Cienaga Momil and Cienaga Sopal. Through a linear regression analysis between field-measured variables and the reflectance values of the Sentinel 2 image, models to represent the spatial variability of the different water quality parameters were obtained. In addition, water quality maps for drinking and aquatic protection life use were created based on Colombian standards. All generated models have a determination coefficient (R2) > 0.7, except for temperature with R2 ≈ 0.6. The five models are characterized by having correspondence in the electromagnetic spectrum with the areas belonging to the near infrared and the visible spectrum. These models, obtained from Sentinel 2 images in combination with GIS tools, allowed a complete analysis that contributes to the spatial–temporal monitoring of the studied aquatic ecosystems.