Global fluvial ecosystems are increasingly impacted by human activities, such as climate warming and land use changes; however, the combined effects of these pressures on river greenhouse gas (GHG) supersaturation and deoxygenation remain poorly understood. This study modeled past global trends (2002-2022) in river GHG saturation, dissolved oxygen (DO) levels, water temperature, and eight other water quality parameters using machine learning models powered by satellite observations. Our findings