Palinspastic reconstructions constrained by sediment geochemistry; a new approach to correlating structurally dismembered lithostratigraphic units in the Caledonides of N. Scandinavia

Lithostratigraphic units become fragmented during continental collisions and these may then undergo different strain and metamorphic histories. Correlating them subsequently can be difficult, especially where primary variations in thickness occur, and even more so if biostratigraphic constraints are poor or lacking. The resulting uncertainties impact attempts to reconstruct the palaeogeography and basin evolution.

As sediment composition is determined by source area composition, weathering before/during erosion, sorting, and biogenic, aeolian and diagenetic/metamorphic additions/alterations, shale sediments derived from the same source area at the same time should have similar chemical characteristics, differentiating them from other sediments. 

Here, we outline part of a regional study of Neoproterozoic to Cambrian shale compositions in the mid- to lower structural levels of the Finnmark Caledonides and parts of the Norbotten Caledonides to test this hypothesis. The aim was to test the validity of presumed correlations between units separated by very large distances in palinspastic restorations. Do similarities in lithostratigraphic sections (crudely, sand vs. mud) reflect anything more than large-scale sea-level variations? Can different source areas be identified?

Major, trace and REE whole-rock data from 98 samples were compared using principal component analysis after the data had been recalculated to centred log-ratio values to mitigate problems associated with the constant-sum effect (Aitcheson 1982). Standard sediment discriminant methods (CIA, MFW and Zr/Sc-Th/Sc plots) support the interpretations given by the principal component analysis but in themselves generally do not show enough differences to yield reliable correlations on their own.

The results confirm some suggested correlations and indicate previously unsuspected ones: Although separated by ~350 km in branch-line/balanced section restorations, the data indicate that the Airoaivi Group in the west of the restored Gaissa Basin (Lower Allochthon) is a correlative of the Vadsø Group in the Autochthon of East Finnmark: The proposal that the Lille Molvik Formation is not part of the Vadsø Group is supported by its chemical similarities with the Tanafjord Group: Inclusion of the Veidnesbotn Formation within the Tanafjord Group, rather than being the basal unit of the Vadso Group, is confirmed by sediment geochemistry. Although these correlations are mostly small-scale and seem localized in importance, they change our overall understanding of the basin evolution, by making some areas that had different sedimentary histories more similar whilst in others they add to the complexity of the basin evolution.

Finally, geochemical differences between the late Precambrian to early Cambrian rocks in the Gaissa Basin of Finnmark and those ~300 km to the south in the Autochthon in Norbotten (Luo Pakte area) reflect deposition from different source areas, despite their detailed lithostratigraphic continuity.

Application of the approach proposed here could usefully be applied to the whole orogen to establish different sedimentary domains in space and time.