<p>Invasion-Percolation (IP) models are used to simulate multiphase flow in porous media across various scales (from pore-scale IP to Macro-IP). Numerous variations of IP models have emerged; here we are interested in simulating gas flow in a water-saturated porous medium. Gas flow in porous media occurs either as a continuous or as a discontinuous flow, depending on the rate of flow and the nature of the porous medium. A particular IP model version may be well suited for predictions in a specific gas flow regime, but not applicable to other regimes. Our research aims to compare various macro-scale versions of IP models existing in the literature and rank their performance in relevant gas flow regimes.</p><p>We test the performance of Macro-IP models on a range of gas-injection rates in water-saturated sand experiments, including both continuous and discontinuous flow regimes. The experimental data is obtained as a time series of images using the light transmission technique. To represent pore-scale heterogeneities of sand, we let each model version run on several random realizations of the initial entry pressure field. As a metric for ranking the models, we introduce a diffused version of the so-called Jaccard index (adapted from image analysis and object recognition). We average this metric over time and over all realizations per model version to evaluate each model&#8217;s overall performance. This metric may also be used to calibrate model parameters such as gas saturation.&#160;</p><p>Our proposed approach evaluates the performance of competing IP model versions in different gas-flow regimes objectively and quantitatively, and thus provides guidance on their applicability under specific conditions. Moreover, our comparison method is not limited to gas-water phase systems in porous media but generalizes to any modelling situation accompanied by spatially and temporally highly resolved data.</p>