As climate change warms rivers, they are losing dissolved oxygen from their water. This process, which is called deoxygenation, was already known to be occurring in large bodies of water, like oceans and lakes. A study that colleagues and I just published in Nature Climate Change shows that it is happening in rivers as well.
We documented this change using a type of artificial intelligence called a deep learning model – specifically, a long short-term memory model – to predict water temperature and oxygen levels. The data that we fed the model included past records of water temperature and oxygen concentrations in rivers, along with past weather data and the features of adjoining land – for example, whether it held cities, farms or forests.
