Quantitative analysis for FluidFlower experiments using multichromatic tracer data, 2023

In the context of geological CO2 storage (GCS), the key question is how reliably CO2 remains in the underground. Laboratory experiments are being conducted at the University of Bergen with the aim of developing and improving simulation models and, at the same time, communicating in a visual way which physical processes are part of them. For this purpose, tracers are used as colour sensors, which allow to visualise the development of the CO2 concentration in the porous material through colour changes on contact. Up until now, the image analysis software DarSIA, which has been developed in Bergen for colour-based analysis, evaluates the images of the experiment only based on monochromatic information by reducing them to a scalar signal at pixel level. However, because multichromatic tracers provide a better representation of the represented concentrations and thus a higher information content, the extension for an image analysis based on multichromatic information should be explored. As a solution for the determination of the represented concentration, this work therefore proposes, among other approaches, an interpolation of defined colour concentration pairs, which can be applied to all colour pixels. In addition to the nearest neighbour method, the kernel method is also tested as an interpolation method. As a result, the kernel method was identified to be a promising approach to determine the concentration of multichromatic tracers for an integration in DarSIA.