Reassessment of paleointensity estimated of a single lava flow from Xitle volcano, Mexico, by means of multispecimen domain-state corrected

Determining paleomagnetic field intensity (paleointensity: PI) for lavas with high reliability and low measurement uncertainty is still difficult to achieve. Some of the factors on which the PI used methods depends are the non-ideal physical and magnetic characteristics of lava sample, grain size, cooling rate effect and thermal stability. Xitle volcano (SW Mexico City) is a good example to illustrate and discuss this problem because several previous PI studies that were carried out on its evolved flow units, have commonly resulted in different mean values with large dispersions. Indeed, 211 published PI data obtained by use of Thellier and microwave experiments gave a mean of 64.1 mu T with a standard deviation of 11.0 mu T. After a careful evaluation, we found that only 134 of these data can be considered reliable, as they meet a set of selection criteria designed in this study. These evaluated data gave an average mean of 62.0 +/- 9.3 mu T. In order to strengthen the PI estimates of Xitle, we conducted a multispecimen domain-state corrected (MSP-DSC) method along one vertical (ca. 4.5 m) and three horizontal (ca. 1.25 m, each) profiles. Top horizontal and vertical profiles have fulfilled a stringent criteria set while central and lower profiles exceeded the alteration check criteria limit and thus are considered unreliable. Accordingly, Xitle PI mean derived from MSP-DSC experiment is calculated at 60.5 +/- 4 mu T, thus in a good agreement with the mean value estimated from previous filtered data. The result and success rate obtained in this study may be ascribed to cooling rate variations commonly found at the lava profile, and indicate that MSP-DSC outcome is governed by the magnetic properties such as the domain-size behavior and the thermal stability of the magnetic carriers present in the treated specimens, as in the conventional Thellier & microwave-style experiments. From these two averages, a combined mean and standard deviation of 61.9 +/- 9 mu T is calculated, which technically is considered the most probable intensity estimate at the Xitle eruption time, ca.370 AD.