Comparison of horizontal-component displacements recorded by an array of 20 receivers due to source two located inside the aquifer, as shown in Fig. 1. The ‘data’ are generated using poroelastic simulations; these are compared to Gassmann elastic (first column; equivalent to the elastic case in Biot's low-frequency regime), and acoustic simulations (second column). The top row shows the differences between the data and Gassmann elastic simulations (first panel), and the data and acoustic simulations (second panel) before injection (BSL), thereby illustrating theoretical limitations due to various modelling approximations. The second row shows differences between the data and Gassmann and acoustic modelling approximations after CO₂ injection (INJ). The third row shows the differences between the seismograms after injection and the baseline, thus highlighting CO₂ signatures in the data and discrepancies from this signature introduced by the Gassmann and acoustic approximations. All plots are normalized, and the numbers in square brackets indicate the ratio between the maximum amplitude of the displayed differences and the maximum amplitude of the corresponding data.