University of Turku



The environmental impacts of fish farming establishments need to be monitored to keep eutrophication at minimum. Currently the data is gathered by external parties and industry gets the results in the form of annual reports. This makes it hard to react to environmental changes when they are actually happening. There is not much data available since the monitoring results are always local and there are also problems in the quality of the data since the monitoring buoys require frequent maintenance. In addition, providing information to internal or external parties is currently slow since the annual reports are the only way to present the data.


Our plan is to create a model in which we merge satellite images and point measurement results into one package to get the best of both worlds. On-site measurements offer the accuracy, while satellite images are able to monitor larger areas and can be used to predict changes. The model utilize machine learning. The results would be shown through an online dashboard, which would contain all the important information with visualisations such as exportable maps and graphs.


We woud do this by integrating Google Earth Engine (GEE) into the dashboard through an application programming interface (API). GEE is a computing platform that allows the user to run geospatial analysis on Google's infrastructure. GEE makes accessing and extracting data easier, and the user can create custom scripts with Javascript and Python.
Bild av projektet