Dissolved oxygen estimation in aquaculture sites using remote sensing and machine learning

Dissolved oxygen (DO) is one of the most critical parameters for aquaculture, as it is vital for all living organisms. The survival, growth and food intake of fish is directly affected by changes in DO concentration. Therefore, the systematic and continuous monitoring of DO is of crucial importance for proper production management. DO does not change the optical properties of water, so it is impossible to estimate its concentration directly from the reflection values of satellite sensors. However, several studies have suggested that it can be estimated indirectly, based on its correlation with other parameters such as chlorophyl-a (chl-a) and sea surface temperature (SST). The present study aims to integrate satellite data, along with in-situ observations to bring forth innovative approaches on how DO can be estimated and monitored on large scale near aquaculture facilities. In this context we exploited daily CMEMS data (chl-a and SST) along with in-situ data (DO) to train a support vector regression (SVR) model. Our in-situ dataset included daily DO measures from Agrilia fish farm in Lesvos for the year 2021. The accuracy of our model was tested using the value of Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE). Our preliminary results indicate that our model performs well locally with promising scalability, which paves the way for the development of real-time monitoring systems for aquaculture.