The key to the discovery lies in the PolandSPAN regional seismic survey, an extraordinary dataset consisting of approximately 2,200 kilometers of onshore seismic reflection profiles acquired with parameters that remain exceptional even by today’s standards. The survey employed a broadband sweep from 2 to 150 hertz, a 960-channel symmetric spread, offsets up to 12 kilometers, record lengths of 12 seconds, tight 25-meter station spacing, and a nominal fold of 480. Compared with classic academic deep seismic surveys such as ECORS across the Pyrenees or POLCRUST in southeastern Poland, which used wider station spacing, lower fold, and shorter offsets, the PolandSPAN acquisition delivered unparalleled images of the entire Phanerozoic sedimentary cover and the underlying crust down to the Mohorovičić discontinuity. It was within this crystal-clear subsurface picture that the high-amplitude reflectivity zones betraying the presence of the deep intrusions first came into focus.

On the seismic sections, the intrusions appear as packages of strong, semi-continuous reflectors, sometimes with a characteristic saucer-shaped geometry, often arranged in step-wise patterns and locally diverging into several separate branches. Such geometries closely resemble seismic images of sill complexes described from the North Sea basin, offshore Norway, and the lower crust of the northern North Sea, as well as field analogues of stepped and lobate sills studied around the world. But a seismic image alone cannot tell you what the intruding rock is made of, how thick it is, or when it was emplaced. Because the features lie far beyond the reach of even the deepest wells in the region, the team had to rely on what they candidly call educated guessing, supported by rigorous quantitative geophysics.

The first step was a careful analysis of vertical seismic resolution. Seismic waves reflected from the top and base of a thin layer interfere with one another, a phenomenon known as tuning. For beds thinner than one quarter of the dominant seismic wavelength, the two reflections cannot be separated, and they merge into a single high-amplitude event. The researchers extracted statistical wavelets from the processed pre-stack time-migrated data within three polygons covering different parts of the intrusion, on profiles PL-5400 and PL-1200. The tuning analysis revealed dominant frequencies of approximately 29 to 30 hertz, with an average of 29 hertz. Assuming P-wave velocities of roughly 6,000 to 6,700 meters per second, typical of mafic igneous rock, this corresponds to a vertical resolution of about 52 to 58 meters according to the classic quarter-wavelength criterion.

With the frequency content established, the team turned to two-dimensional seismic forward modeling to test competing lithological scenarios. They built a 160-kilometer-long, 15-kilometer-deep seismic-geological model based on interpreted profile PL-5400, assigning velocities and densities to each sedimentary layer using data from the Pasłęk IG-1 and Prabuty IG-1 deep research wells, and applying a vertical velocity gradient in the Precambrian basement constrained by well data and deep seismic refraction studies. Densities were estimated with the Gardner equation. Three candidate lithologies were tested for the intrusions: granite-granodiorite with a velocity of about 5,950 meters per second and a density of 2.67 grams per cubic centimeter, basalt-like parameters of about 6,240 meters per second and 2.76 grams per cubic centimeter, and dolerite with about 6,700 meters per second and 2.87 grams per cubic centimeter.

The results were striking. In the granite-granodiorite scenario, the acoustic impedance contrast with the host basement was so weak that the modelled intrusion was almost invisible, producing synthetic amplitudes far lower than those observed in the real seismic data. The basalt scenario fared somewhat better but still generated a low-amplitude response inconsistent with the observations. Only the dolerite scenario, with its significant velocity contrast and density difference exceeding five percent of background values, produced a high-amplitude synthetic image closely matching the real data. Combined with the fact that numerous wells in the Baltic Basin, including the Pasłęk IG-1 deep well that calibrates profiles PL-5400 and PL-1200, have documented lower Carboniferous dolerite intrusions in shallower levels, the conclusion was clear: the deep bodies are most probably doleritic.

To estimate thickness, the team constructed a wedge model simulating a dolerite layer thinning from 200 meters to zero within the crystalline basement, computed with a full-wave algorithm and a zero-phase 29-hertz Ricker wavelet. The model showed that reflections from the top and base of the wedge can be separated only when the layer is thicker than about 115 meters; as the wedge thins, constructive interference builds to a maximum at the tuning thickness of roughly 57 to 58 meters, below which amplitudes decay and the two reflections become indistinguishable. A direct comparison between a real seismic trace from profile PL-5400 and a synthetic trace extracted from the wedge model at an equivalent thickness of about 58 meters showed a close correspondence in peak-trough separation, relative amplitudes, and overall wavelet shape, validating the entire approach.

Armed with these quantitative constraints, the researchers interpreted the intrusions in detail. The top of the intrusion produces a positive-amplitude reflection, while the base yields a negative one, and the amplitude of both varies systematically with thickness. Where amplitudes are strongly tuned, the intrusion thins to around 60 meters; where top and base are clearly separated with internal reflectivity between them, thickness exceeds 120 meters and locally reaches 200 meters or more. Polarity reversals and abrupt amplitude cut-offs mark zones of significant lateral thinning, tapering, or termination, and step-wise geometries are clearly visible, although their precise origin at such depths remains an open question. The authors caution that compositional heterogeneity, internal layering, or variations in the host rock could also contribute to the observed seismic response and warrant further integrated study.

Dating the intrusions relied on the rich regional geological record of the East European Craton. Four magmatic episodes are known from the area: Mesoproterozoic anorogenic activity around 1.54 to 1.45 billion years ago that formed the anorthosite-mangerite-charnockite-granite suite, late Ediacaran flood basalts of the Volhynia Large Igneous Province around 580 to 545 million years ago, Mississippian alkaline magmatism of the recently recognized Lublin-Baltic Igneous Province dated to about 352 to 344 million years ago, and latest Carboniferous to early Permian magmatism. The deep intrusions cross-cut the ancient basement rocks, ruling out a Proterozoic origin, while the nearest known products of Ediacaran and Permo-Carboniferous magmatism lie at least 200 kilometers away. Similar extensive intrusions imaged by PolandSPAN data within the Silurian to Lower Devonian sedimentary cover of the Lublin Basin are known to be early Carboniferous in age. Together with the close spatial association with shallow Mississippian dolerite sills and alkaline massifs drilled in the Baltic Basin and the nearby Mazury High, the evidence points unambiguously to a Mississippian, or early Carboniferous, age.

Beyond illuminating a previously hidden chapter of European magmatism, the study delivers a methodological blueprint with global reach. By combining quantitative assessment of seismic resolution and tuning thickness with two-dimensional forward modeling anchored in regional geological constraints, geoscientists can now estimate the lithology, thickness, and geometry of deep igneous bodies that will never be sampled by drilling. Such techniques have already proven their worth in studies of sill complexes in the North Sea and offshore Norway, and the Polish team’s work extends them into the deep crystalline basement of an ancient craton. As the authors note, the approach inevitably rests on assumptions and should be applied with appropriate caution, but it transforms educated guessing into a disciplined, testable inference, offering a powerful new window on the hidden plumbing of the Earth’s crust.

Subject of Research: Lower Carboniferous igneous intrusions in the crystalline basement of the Baltic Basin, Poland, characterized through seismic data interpretation and forward modeling

Article Title: Lower Carboniferous igneous intrusions within the crystalline basement of the Baltic Basin (SW edge of the East European Craton, Poland) – insight based on seismic data interpretation and seismic forward modelling

Article References: Krzywiec, P., Słonka, Ł., & Poprawa, P. (2026). Lower Carboniferous igneous intrusions within the crystalline basement of the Baltic Basin (SW edge of the East European Craton, Poland) – insight based on seismic data interpretation and seismic forward modelling. Solid Earth, 17(6), 825-853. https://doi.org/10.5194/se-17-825-2026

Image Credits: AI Generated

DOI: 10.5194/se-17-825-2026

Keywords: Baltic Basin, igneous intrusions, dolerite sills, seismic reflection, seismic forward modelling, East European Craton, Lublin-Baltic Igneous Province, Mississippian magmatism, seismic tuning, crystalline basement, PolandSPAN, geophysics