Monitoring the eutrophication using Landsat 8 in the Boka Kotorska Bay

Abstract

         This study proposes a methodology for monitoring concentrations of chlorophyll a (Chl-a) and the state of eutrophics in small bays or in the immediate vicinity of the coast. This kind of monitoring is of interest since such areas have not been addressed well using the usual satellite methods (such as MODIS) due to inadequate spatial resolution. We present an estimation approach for Chl-a concentration based on Landsat 8 (L8) satellite images using the ground truth (GT) data for the day of overflight. Additionally, two classifiers (daily and yearly) of the state of eutrophication, that use the Chl-a estimated values, are presented. The accuracy of the proposed method is evaluated using the leave-one-out cross validation, and it is within limits theoretically expected of an L8-based approach. The results from the classifiers are compared with the GT data and it is shown that daily classifier is able to classify the area of interest with an incidence of false positives less than 2%.