Dissolved oxygen is a dynamic parameter that is influenced by several factors including temperature, wind, and biological activity. In aquaculture pens, it is essential to monitor dissolved oxygen concentration, since it impacts the health, growth, and feeding activity of fish. The purpose of this project was to develop a deep learning model capable of accurately predicting dissolved oxygen concentration in fish pens up to 24 hours in advance. This work supports better decision-making to reduce the energy use, costs, and environmental impact of aquaculture.
