Développement d'un outil chronostratigraphique pour les archives climatiques : datations absolues (K/Ar,⁴⁰Ar/³⁹Ar) et paléomagnétisme appliqués aux laves

The understanding of climatic mechanisms and rapid climate changes requires a high-resolution, robust, and precise timescale which allows long-distance and multi-archives correlations.An appropriate tool to construct such a timescale is provided by the Earth magnetic field (EMF). The EMF is independent from climatic variations and its past evolution, global at the surface of the Earth, is recorded by most of the geological/climatic archives. Sedimentary sequences provide continuous records of relative intensities of the EMF on timescales usually based on ice core age models or orbital tuning. Lavas, though discontinuously emitted through time, record the absolute intensity of the EMF during their cooling at the surface of the Earth. Lavas are dated using 2 complementary methods: ⁴⁰Ar/³⁹Ar and K-Ar, both independent from climatic parameters. Lavas have therefore the potential to deliver tie-points (age-paleointensity couples) enabling the time calibration of sedimentary sequences and their transfer onto absolute intensity scale and chronological time scale. This timescale can then be transferred to other climatic archives. The present study focusses on the last 200 ka with lavas sampled from young volcanoes of Ardèche (South Massif Central, France) and recent phases of volcanism in the Canary Islands.Lava flows from Ardèche provided unexploitable paleointensity results and ages with large uncertainties. Therefore, they failed to provide suitable tie-points. However, our geochronological results evidence how crucial the combination of both the K-Ar and 40Ar/39Ar methods is to test the accuracy and geological meaning of the ages. Ardèche lavas have abundant mantellic and crustal xenoliths, potential carriers of excess ⁴⁰Ar*. Our study suggests that the argon excess is located in sites that decrepitate at low temperature (<600°C). Because ⁴⁰Ar/³⁹Ar ages are not affected by excess ⁴⁰Ar*, they provide reliable results. The new age dataset indicates that the volcanic activity of Ardèche can be divided in 3 phases: the oldest one (180±30 ka) took place in the northern part of the studied area and 2 younger phases are expressed in the South (31±4 ka and 24±8 ka).The study of the Canarian lavas produced 14 tie-points (9 out of 14 dated combining K-Ar and ⁴⁰Ar/³⁹Ar results). These data have been added to the available ones for the same time period. The published data have been selected on the basis of robust analytical protocols and accuracy. The 51 data finally selected are compared to available sedimentary stacks. Over the last 80 ka, the volcanic data corroborate the calibration of GLOPIS-75, initially based on volcanic and archeomagnetic data between 10-20 ka and the low intensity observed in the Laschamp excursion. Three newly produced data, dated between 45 and 60 ka, extend the database initially used to older periods and they are also consistent with the initial calibration of GLOPIS-75. Between 80 and 140 ka, though volcanic data have significant uncertainties (in age and/or paleointensity), they are consistent with available sedimentary records and validate their calibration level on the long-term. At a shorter time scale, volcanic data corroborate the intensity low reached during the older phase of the Blake excursion (120 ka) by PISO-1500, whereas this low does not appear in SINT-2000. For ages older than 140 ka, not only the volcanic data are scattered, but also the sedimentary records are different from one another and no conclusions could be drawn. Finally, 2 of our data suggest a brief geomagnetic event around 155 ka. Such an event cannot be seen on available global sedimentary stacks or models, even though some individual studies report a local geomagnetic event around 150 ka (Austria, Russia, and China Sea).Développer une échelle de temps à haute résolution temporelle et commune aux différentes archives climatiques est une étape importante afin de quantifier avec précision les rapides variations climatiques passées et pour les placer dans un cadre chronologique unifié facilitant leurs inter-comparaisons et la quantification d’éventuels déphasages entre évènements, marqueurs ou archives climatiques.Le champ magnétique terrestre (CMT) regroupe l’ensemble des caractéristiques désirées pour développer un tel outil chronostratigraphique (expression dipolaire globale à la surface du globe, enregistrement dans diverses archives, variations en intensité indépendantes des variables climatiques). Bien que porteurs d’enregistrements continus, les sédiments ne donnent accès qu’aux variations relatives d’intensité du CMT. De plus, quand leur échelle de temps ne peut plus être placée sur celle des glaces polaires, elle est généralement obtenue par forçage orbital. Les laves, émises sporadiquement, enregistrent l’intensité absolue du CMT et sont datables par méthodes ⁴⁰Ar/³⁹Ar et K-Ar (indépendantes des variables climatiques). Elles fournissent ainsi des couples âge-paléointensité (A-PI) permettant de calibrer les enregistrements sédimentaires et de les transférer sur des échelles de temps et d’intensité absolues. L’échelle de temps ainsi obtenue est par la suite transférable à diverses archives climatiques. Cette étude se focalise sur les derniers 200 ka. Les laves étudiées proviennent des jeunes volcans d’Ardèche et des phases récentes du volcanisme canarien. Les laves ardéchoises ont délivré des résultats de paléointensité non exploitables et des incertitudes trop importantes sur les âges. Aucun couple A-PI pertinent n’a donc été obtenu. Cependant, nos résultats géochronologiques démontrent l’importance de combiner les 2 méthodes de datation K-Ar et ⁴⁰Ar/³⁹Ar pour tester l’exactitude et la signification géologique des âges obtenus. Pour ces laves, porteuses d’indices de contamination crustale et mantellique, nous suggérons que l’excès d’argon est situé dans des sites de rétention basses températures (