Temporal Prediction of Net Ecosystem Exchange (NEE) by Transformer Model

This research aims to predict Net Ecosystem Exchange (NEE) across temporal and spatial dimensions using remote sensing, climate, and eddy-covariance (EC) flux datasets. The study focuses on 185 forest and woodland FLUXNET sites globally, analyzing the response of disturbance and climate variations on Gross Primary Productivity (GPP) and Net Ecosystem Exchange (NEE) over time. The research involves data preprocessing, feature engineering, model selection, and training to predict NEE accurately. The results are visualized through graphs and charts, providing insights into the patterns and trends of NEE over time and space. The study is compelling and impactful as it contributes to the understanding of carbon sequestration through the fast cycle of photosynthesis and helps reduce total CO2 in the ecosystem. Additionally, the data produced by the models can be used in new and ongoing research to understand and mitigate the effects of global climate change.