Using Landsat and in situ data to map turbidity as a proxy of cyanobacteria in a hypereutrophic Mediterranean reservoir

[object Object]Satellite remote estimates of phycocyanin (PC) have become valuable for monitoring the quality of inland waters affected by harmful cyanobacterial blooms. In this study, we developed an algorithm for mapping turbidity as a proxy of PC content through Landsat 8 Operational Land Imager (OLI) data and in situ measurements. The chosen study site is Karaoun Reservoir, in Lebanon, a hypereutrophic freshwater body where turbidity is mostly driven by cyanobacteria. Satellite images were corrected for atmospheric effects with the 6S (Second Simulation of the Satellite Signal in the Solar Spectrum) code which proved to be more accurate than the DOS (Dark Object Subtraction) approach with R = 0.98 and R = 0.5, respectively. A strong relationship was found between turbidity and PC measurements (R = 0.92, R2 = 0.86), as well as between turbidity and the ratio of band 5 to band 4 of the OLI (R = 0.88, R2 = 0.77). Results reveal a promising performance of the algorithm for predicting PC concentrations with high correlations determined through simple linear regression analysis for both the calibration (R = 0.92, R2 = 0.85) and validation (R = 0.88, R2 = 0.78) periods. An application of the approach to a set of historical Landsat images revealed a time series of cyanobacterial bloom occurrence with high variation in surface area at the study site. The algorithm is considered to be suitable for retrieving cyanobacteria in highly eutrophic waters dominated by cyanobacteria where turbidity is mostly a function of the latter. This approach will improve monitoring cyanobacterial blooms on a spatial and timely basis.