Analysing lithological complexity in outcrop-scale seismic models of interbedded siliciclastics and carbonates from Svalbard

Thin sedimentary units typically constitute a large part of the volume of reservoirs targeted for hydrocarbons, CO₂ storage, or groundwater but are considered to be below resolution thresholds of conventional seismic reflection data. Nonetheless, they affect seismic imaging through interference, and where lithology varies significantly over short distances, reflection characteristics may change laterally. This is especially valid for interbedded siliciclastic-carbonate systems characterised by lateral and vertical changes in lithology. We use high-resolution digital outcrop models of two interbedded siliciclastic-carbonate outcrops from Svalbard as input for seismic modelling using a 2D point-spread-function based convolution. The digital outcrop models show Carboniferous to Triassic strata, from southern Spitsbergen and Bjørnøya. We investigate detection thresholds (bed thickness, layer geometry, lithologic variability) and discuss their implications for resolvability on seismic data. Our base-case seismic models illustrate that units below the vertical resolution can be detected, even at high velocities (5–5.5 km/s). We present 6 scenarios to test the influence of thin (metre-scale) beds, lithological variability, faults, karstification, and imaging parameters on the seismic models. They illustrate that in mixed successions variations in reflection amplitudes, dimming and tuning effects become important for recognising thin units. Additionally, thin interfingering units can create artefacts such as apparent “faults”, thus leading to misinterpretation of seismic sections in this type of depositional setting. Varying seismic imaging parameters, such as adding random noise or reducing dominant frequency decrease the detectability of thin units.