Caracterización paleomagnética de procesos deformacionales en cuencas intraplaca (alto Atlas central)diapirismo, compresión e intrusiones ígneas

Resumen

The Atlas System comprises a set of intraplate basins inverted during the Cenozoic as a consequence of the Africa-Iberia-Eurasia convergence. These basins were developed during the Mesozoic in two rifting stages (Triassic and Early Jurassic), influenced both by the opening of the Atlantic and the evolution of the Tethys. The study area is located in the middle of the Central High Atlas (Morocco) and is characterized by NE-SW to ENE-WSW tight anticlines separating open, gentle synclines. The Mesozoic sequence reaches big thickness (more than 6-7 km) in the depocenters. This sequence is mainly composed by Lower-Middle Jurassic marine carbonates that pass upwards into continental red beds (Middle Jurassic sediments). No younger sediments crop out in the study area but in the northern basins Upper Jurassic to Lower Cretaceous continental sediments can be observed. Basalts linked with the CAMP event and emplaced in sub-aerean conditions at the end of the Triassic crop out in the core of the anticlines, together with Upper Jurassic plutonic rocks (mainly troctolites, gabbros and sienites) and Triassic shales and salts. Associated with the Jurassic volcanic event several dykes and, in the northern basins, two sets of basaltic lava flows that were emplaced during the Late Jurassic and the Early Cretaceous can be observed. Several aspects (growth strata, angular unconformities in the limbs of the anticlines, etc.) evidence a major halokinetic activity during the Early-Middle Jurassic, which conditioned the sedimentation in the region defining areas with differential subsidence. Besides, this halokinetic activity deformed the Jurassic carbonates during the development of diapirs and salt-walls. The Jurassic carbonates of the Central High Atlas record a widespread chemical remagnetization, carried by stable single domain (SSD) magnetite, which has been previously dated as Albian-Cenomanian (ca. 100 Ma). This is an interfolding remagnetization since it is temporally bracketed between two folding stages (Jurassic extension and Cenozoic compression). Working with this kind of remagnetizations, the small circle (SC) tools allow (i) to calculate the remagnetization direction and (ii) to restore the beds at the remagnetization time i.e., to calculate the paleodip of the beds. Once that the paleodips are know, they can be used to restore partially the structure and to show how was the structure during the remagnetization acquisition. In regard with the methodological development of the SC tools, in this doctoral thesis is presented a new software (PySCu) allowing to perform the necessary SC calculations in a simple way. As PySCu is written in language Python, the application is open and cross-platform. Moreover, this incorporates an improvement in the calculation of the uncertainty associated to the calculated remagnetization direction: using bootstrap the program estimates the error propagation in the calculated direction resulting from the uncertainty of the mean paleomagnetic direction and the bedding. Using the SC tools, three main starting hypotheses must be considered: (i) the remagnetization must be geologically synchronous, (ii) there is an absence of vertical axis rotation (VAR) and (iii) the pre- and post-remagnetization folding are coaxial. Through simulations of artificial data that consider both presence of VAR and non-coaxiality the effect in the results of the non fulfillment of the starting hypotheses was evaluated. The presence of VAR affects strongly the calculation of the remagnetization direction and systematically higher inclinations than expected are obtained. On the other hand, VARs generate characteristic patterns in the calculation of the paleodip, over- and under estimating symmetrically the paleodip in opposite limbs. These features can help to detect the presence of VARs. On the other hand, the non-coaxiality affects in lesser degree both to the remagnetization direction and the paleodip calculations. The comparison between simulations and real data shows that the study area fulfills the starting hypotheses necessary for the application of the SC tools. Moreover, the comparison of the bedding between sites affected mainly by the pre-remagnetization deformation with sites in which the post-remagnetization deformation dominates, indicates that both folding states are nearly coaxial. Present-day and restored (at the remagnetization time, ca. 100 Ma) geologic cross-section have been compared and this practice has shown to be a useful tool to unravel and clarify some geological aspects present in the study area: (i) the region is affected by a regional cleavage considered alternatively as Jurassic and Cenozoic by different authors. The folds with axial-plane cleavage clearly post-date the remagnetization because they completely unfold after the partial restoration. This means that cleavage developed during the Cenozoic compression. (ii) On the other hand, the restoration of several anticlines shows a different degree of development of the structures at the remagnetization time. Some of them were already structured at ca. 100 Ma, with steep limbs that become progressively horizontal towards the synclines, whereas in other structures the pre-100 Ma structure is gentler and limited to the proximity of their cores (affecting only to the nearest 500 m). This pre-remagnetization deformation is mainly the consequence of the halokinetic process that generated diapirs and salt-walls with different degrees of development. These structures seem to be related to basement normal faults that controlled the tectonic subsidence. Besides, the structure inherited from the extensional stage had a main role controlling the subsequent deformation during compression. As well as the beds can be restored, other structural elements that can be related to bedding can also be restored using the paleodip, as for example the petrofabric or the magnetic fabrics. It is a common procedure to compare the in situ magnetic fabrics with the totally restored (rotating the bedding to the horizontal) ones; knowing the paleodip, they also can be compared with the partially restored (rotating the bedding to the paleobedding) magnetic fabrics. The restoration of the anisotropy of the anhysteretic remanent magnetization (AARM) has allowed to put forward significant advances in the understanding of how the SSD magnetite grains that carry the remagnetization grow. The AARM has been measured in sites with different dips and paleodips, and the orientation of the principal axes of all sites has been compared before apply any bedding correction, after apply the total bedding correction (i.e., restoring the bedding to the horizontal) and after apply the partial bedding correction (i.e., restoring the bedding to their attitude at the remagnetization time). Comparing these three stages, the best clustering of the principal axes is reached after the partial bedding correction, when a prolate ellipsoid with horizontal magnetic foliation and horizontal NNW-SSE magnetic lineation, parallel to the general tectonic extension is defined. This has been interpreted because either: (i) the neoformed magnetite grains grew without following any preexisting structure or, conversely, (ii) the magnetite grains grew following the extensional tectonic constraints present in the Atlas at the remagnetization time, producing a weak but well-defined magnetic fabric. Furthermore, the absence of compressive fabrics in the AARM shows that the magnetite grains were not affected by the Cenozoic cleavage present in the study area. (iii) We interpret these facts as an indicator that the magnetite grains grew replacing pyrite crystals that, on one hand, offers an isotropic frame within which the grains can grow following the tectonic constraints and, on the other hand, the pyrite grains deflect the deformation directions associated with cleavage development. The analysis of the anisotropy of the magnetic susceptibility measured at room temperature (RT-AMS) and its comparison with different subfabrics (LT-AMS and AARM) allow to unravel four different behaviors in RT-AMS. Type 1 RT-AMS shows the same behavior that the already described by AARM and RT-AMS is not coincident with LT-AMS. Conversely, type 3 and 4 RT-AMS are coincident with the LT-AMS but not with AARM; in these cases the magnetic lineation is parallel to the intersection lineation defined between cleavage and bedding (it is a tectonic compressive fabric); the difference between the two fabrics is that the magnetic foliation in type 3 is parallel to bedding and in type 4 is parallel to cleavage. Finally, type 2 shows intermediate behavior between the previous ones. The comparison between the different subfabrics, as well as the interpretation of rock magnetic measurements, allow to explain the type 1 RT-AMS as carried by superparamagnetic magnetite neoformed during the remagnetization process; these grains have the same distribution than the SSD magnetites that carry the remagnetization. On the other hand, type 3 and 4 RT-AMS are carried by phyllosilicates, which at difference with ferrimagnetic grains are affected by the Cenozoic compression. The Jurassic gabbros that crop out at the core of the anticlines have also been analyzed paleomagnetically. A stable paleomagnetic component with maximum unblocking temperatures around 580ºC that is carried by magnetite has been isolated. The mean paleomagnetic directions show a strong scatter when all structures are analyzed together. This scattering has been interpreted as a consequence of tectonic movements. However, when the mean paleomagnetic directions are plotted for each structure separately, the directions are located over same small circles, with horizontal axes parallel to the main trend of each respective structure; this is evident in the western structures, where the scatter is lower and there are available more site-mean paleomagnetic directions. The dispersion of the paleomagnetic directions over small circles parallel to each structure is interpreted as the result of the Cenozoic folding. Besides, most of the sites record counter-clockwise rotation (looking towards the NE / ENE) that means that the structures present a dominant vergence towards the NW / NNW. The magnetic fabrics in the gabbros are coincident with the petrofabric defined by the plagioclase crystals. Before any correction, the orientation of the main axes show a strong scattering between sites. However, and using the information that come from the paleomagnetic analysis, when the magnetic fabrics are restored to the emplacement position, they show a better cluster and are coherent in each structure. According to paleomagnetic results and magnetic fabrics two kind of structures have been differentiated: (i) in the western structures, where the ratio between ductile and igneous rocks is higher, the recorded rotations are bigger. The magnetic fabrics show mainly horizontal magnetic foliations and horizontal NW-SE magnetic lineations that we interpret as the reflect of the regional tectonic extension during the intrusion of the igneous rocks. (ii) On the other hand, in the eastern structures the ratio is lower and the cores of the anticlines are mainly composed by igneous rocks. In these cases the rotations are smaller and the magnetic fabrics show more complex patterns; the magnetic lineations are mainly horizontal and N-S to NE-SO, and the magnetic foliation presents different attitude. This is likewise consequence of the emplacement of the magma with limited space and the magnetic fabrics reflect the different flux controlled by the structural frame.