| Abstract: | When pyroclastic deposits settle at high temperature in a terrestrial magnetic field they memorize their setting temperature. This temperature varies from one type of deposit to the other, from ambient temperature for some airfall ashes up to hundreds of degrees centigrade for nuées ardentes or pumice flows. These temperatures also vary significantly with distance from the eruptive vent. At old events, it is often very difficult first to establish the nature of the deposit because of crystallizations or cements created by diagenesis, and then to locate the eroded eruptive centers. The exact assessment of the setting temperature allows one, on the one hand, to find both the nature and the emplacement process of the deposit and, on the other hand, to find the location of the eruptive centers, by working out a temperature map.The methods used up to now are all based on the analysis of objects buried in the flow, or on that of the surrounding rocks. Where burnt wood or heated objects are concerned, this analysis may be chemical or based on fossil thermoremanent magnetization in the case of potteries or rocks. Such objects are not always to be found, so the method proposed here uses samples drawn from the deposit itself. The principle, derived from Thellier's method, compares the natural magnetization lost during heating to the loss of total thermoremanent magnetization acquired during cooling in a known field. This method essentially differs from that of Thellier by a direct reading of the magnetization left at any high temperature. This process, which requires only two heatings, enables one to avoid any difficulty linked to magnetic interactions and to lessen the effect of mineralogical transformations. |
