This project focuses on improving the understanding of vegetation–climate interactions by combining machine learning techniques with existing ecosystem modeling approaches. Traditional process-based vegetation models rely heavily on observational data to calibrate their parameters, but the limited availability of such data makes accurate modeling challenging. Machine learning provides an alternative approach by identifying patterns in complex climate and vegetation datasets and generating predictive models without requiring predefined relationships between variables. In collaboration with Dr. Bi, our team processes and prepares vegetation and climate datasets to develop machine learning models that predict vegetation characteristics based on climate conditions. The project utilizes vegetation data from the MODIS (Moderate Resolution Imaging Spectroradiometer) satellite dataset and climate data from the gridMET dataset. Two primary modeling tasks are addressed. The first involves predicting vegetation distribution using either observational datasets or data generated by dynamic vegetation models within Earth System Models (ESMs). The second focuses on predicting a vegetation structural indicator, specifically Leaf Area Index (LAI), using observational data or phenology model outputs. The machine learning models are evaluated based on their predictive performance under current climate conditions as well as their ability to generalize across different climate scenarios. This assessment helps determine which models are most suitable for further analysis. The project deliverables include processed climate and vegetation datasets, machine learning models for predicting vegetation distribution, LAI, and SAI, and supporting code for model training and evaluation. Documentation will also describe the data processing workflow, model development process, dataset formats, and instructions for using the developed software tools.