Penn State will improve subsurface characterization and quantification by data fusion. This will include surveying the literature of available technologies for subsurface characterization of CO2 in various types of reservoirs, design controlled laboratory experiments to collect data using different sensors (4D X-ray microtomography, acoustic emission, stress and deformation, core-scale flow rate) and using Machine Learning techniques to automate data assimilation, visualize the combined datasets, exploit multi-sensory data, and classify the images into meaningful categories. This will allow for a combine physics-based and Machine Learning techniques to upscale CO2 saturation from pore/core to the target scale.