Paleomagnetic dating of ferricretes in New Caledonia: constraints on the morphogenesis and paleoweathering of the Grande Terre

Although the description of the emplacement and the weathering of the New Caledonia peridotites are well documented in the literature (Trescases, 1975), the processes and the age of the planation surfaces formed upon the ultramafic massifs are poorly documented. Ambiguities remain on: 1) the timing and modality of weathering, 2) formation of supergene nickel ore and 3) the role of the parameters controlling its evolution: •The overall geometry of the ophiolitic nappes and structures possibly inherited from the obduction history, •The lithology of the bedrock (dunite, wehrlite, harzburgite, gabbro, amount of serpentinisation ...), •The post-obduction uplift history, and the paleogeographic and paleoclimatic evolution of New Caledonia up to Present. •The geomorphologic evolution related to recent tectonics and sea level changes, Several planation surfaces have been recognised along the island and their correlation in the southern part of the Grande Terre has been done (Chardon and Chevillote, 2006), but no reliable ages could be attributed up to now. Authors established relationships with recent tectonic events; however several issues remain unresolved: •Are all planation surfaces of the same age? Are they deformed (faulted) at a regional scale? •Or alternatively are there stepped planation surfaces of different ages? Are differences in elevation related to variations of the uplift rate? •Are there 1, 2 or more generations of weathering profiles associated to those surfaces? Are possible differences in petrography, mineralogy and geochemistry related to different ages and thus to contrasting paleoclimatic and/or geodynamic settings? The weathering being still active on the peridotites in the current differentiated landscape and tropical climate, the beginning of the lateritic profiles development is not well constrained and assumed to have occurred between 34 Ma (end of obduction) or 25 Ma (last granodiorite intrusions) and Present. Llorca & Monchoux (1991) have found hollandite minerals which are judged as reliable to estimate the age of weathering (Vasconcelos et al., 1992). Work is in progress to analyse this kind of material. Moreover, ferricretes and various ferruginous materials are largely present in the weathering profiles and in the fluvio-lacustrine sediments as well. They would record the ancient geomagnetic field providing means of age determination (e.g. Théveniaut & Freyssinet, 2002 ; Ricordel, 2007). In tropical soils, most of the primary remanence carrying minerals are dissolved during weathering and secondary magnetic minerals, such as goethite and haematite, are formed in situ acquiring a crystallisation (or chemical) remanent magnetization (CRM). Assuming that no significant tectonic movements occurred between New Caledonia and Australia during the Neogene, the paleomagnetic pole recovered by demagnetizing the CRMs are plotted on the apparent polar wandering (APWP) reference curve of the Australian Plate, providing an age for the different parts of the paleoweathering profiles (ie ferricretes capping the lateritic profiles, and iron oxide concretions occurring deeper in the profiles). The data and interpretations we present here are based on paleomagnetic analysis of ferricretes capping the weathering profiles along the Grande-Terre. Well constrained ages have been obtained for the first time. Four sub-sites along a section crosscutting the Tiebaghi plateau (Northwest) have consistently provided an age of ca. 25 ± 5 Ma. In contrast, the samples of the Goro site (South) span over a larger time range of 0-5 Ma to 25 Ma, highlighting possible stepped paleosurfaces. These paleomagnetic data give the first dating constraining the morphogenesis of the Grande Terre paleolandscapes. They could allow reliable correlation of the planation surfaces. The planed future developments include: •The study and dating by paleomagnetism and U-Th/He geochronology (Lippolt et al., 1998) of a larger number of lateritic ferricretes and an analysis of fluvio-lacustrine ferruginous material (iron oxides concretions), which are of major importance for discriminating the different paleosurfaces, refining the steps of downwards progression of weathering, and reconstructing the regolith history, •A comparison with isotopic curves and paleoclimatic evolutions at a global and a regional scale in the Southwest Pacific during the Neogene, •Additional O and H isotopic data on Fe oxides from the same ferricretes, iron oxide concretions and from kaolinite sampled in weathering profiles developed from felsic rocks below similar planation surfaces, to precise the paleoclimatic conditions of the weathering (Girard et al., 2000), •Additional thermochronological data (Gallagher et al., 1998) and petrographic studies of the weathering profiles from the ferricrete until the unweathered bedrock. They all should help to decipher the roles of climate, tectonics and lithology respectively