High-resolution seismic imaging of the Koillismaa Layered Igneous Complex, Northern Finland

The demand for raw materials has scaled exponentially in recent times due to their applications in various areas, including finished goods, energy, electronics, and lithium-ion batteries. The Koillismaa Layered Igneous Complex (KLIC) in northern Finland holds great potential to host several critical raw materials such as cobalt, nickel, PGEs, etc. It is a mafic-ultramafic complex spanning over a distance of ~50-60 km and is linked by a high gravity and magnetic anomaly. Drilling in the area confirmed the presence of ultramafic rocks at a depth of ~1.4 km from the surface. Extensive petrophysical and lab studies were conducted, and a preliminary Common Earth Model (CEM) was made mainly based on the potential field data inversion with constraints from the borehole. Two regional 2D seismic profiles were acquired under the ERA-MIN 3 sponsored SEEMS DEEP Project (2022-2025) with the aim to map the regional seismic reflectivity in the area and to constrain the geometrical architecture of the KLIC. The processing of the 2D seismic data followed the standard workflow, e.g., dip-moveout followed by the post-stack time migration with constant velocity. This procedure is effective for simple geological settings, i.e., with gentle dips. In the case of KLIC, the subsurface geology is structurally complex; therefore a transition from the standard time-domain imaging to the depth-domain imaging is required. One of the main challenge in doing this is the unavailability of a robust velocity model. We used first-arrival traveltime tomography (FATT) and acoustic Full Waveform Inversion (FWI) to build the high-resolution velocity model. We used the steepest-descent optimization algorithm, optimal-transport objective function, and inverted for the P-waves only using the vertical-component data. The depth details for the FATT-derived velocity model were limited to a few tens of meters from the surface as compared to ~1 km for the FWI-derived velocity model. For migration, we performed prestack Kirchhoff depth migration (KPreSDM) with both FATT- and FWI-derived velocity models. In the latter case, we did not observe much uplift in terms of the overall imaging, partially may be due to the lesser sensitivity of the ray-based KPreSDM towards the input velocity model and limited velocity details with depth. Therefore, we also tested least-square KPreSDM to obtain the depth image with better amplitude fidelity. We then tested wave-equation based Reverse Time Migration (RTM) due to its ability to better handle the complex media using both the derived velocity models. RTM with FWI-derived velocity model provided us with the best imaging overall until a depth of ~5-6 km, establishing the merit of these advanced methods for high-resolution seismic imaging of the geologically complex settings such as the KLIC. The obtained results showed good correlation with the available petrophysical data, observed gravity & magnetic highs, available CEM, and the controlled source electromagnetics-derived resistivity model, which was also acquired during the SEESM DEEP project. Overthrusting with regional-scale faults was imaged, and a funnel-shaped geometry of the KLIC was established. Seismic imaging also suggested a more structural or compositional heterogeneity within the mafic-ultramafic KLIC body.