Global warming and climate change have significantly altered vegetation ecosystems, affecting how plants perform photosynthesis. Monitoring these biological processes is crucial for understanding the Earth’s carbon cycle and informing climate policies, yet accurately tracking them remains challenging due to the large spatiotemporal variability in vegetation responses to external factors. This study hypothesizes that incorporating environmental factors can effectively represent changes in carbon exchange processes.
   In this research, we use machine learning and deep learning models to estimate gross primary productivity (GPP), ecosystem respiration (Reco), and net ecosystem exchange (NEE) based on various input features, analyzing which variables most impact plant carbon exchange. We are conducting two separate studies utilizing data from polar-orbiting and geostationary satellites to leverage their respective spatial and temporal advantages in monitoring terrestrial carbon flux. The models are applied and validated using ground-based measurements from Eddy Covariance Flux Towers. This approach demonstrates the potential of satellite observations combined with environmental variables to analyze how external factors influence GPP variations.