Palaeomagnetic dating of two recent lava flows from Ceboruco volcano, western Mexico

Two lava flows from the Ceboruco volcano in west-central Mexico were sampled for palaeomagnetic dating. The younger one was emitted in 1870 and used to validate the method, while the older one known as Ceboruco flow is of unknown age but probably younger than similar to 1005 AD and older than 1528 AD. Each flow was sampled in at least four sites, in order to unravel between site variations. For the 1870 flow, between site differences were notable and additionally post-cooling block movements were important; therefore, two sites had to be rejected. Three sites from the vent area and one at the tip of the 1870 flow provided well-constrained directions. This is also true for Ceboruco lava flow, and overall mean directions and palaeointensities were then used for palaeomagnetic dating applying the Matlab tool archaeo_dating and the global palaeosecular variation model SHA.DIF.14k. For the 1870 lava flow, the dating resulted in an age ranging between 1755 and 1871 AD (95 per cent probability level), which includes the real emplacement age. In addition, the Ceboruco lava flow was dated between 1000 and 1134 AD, which is close to the large plinian Jala eruption producing the crater of Ceboruco volcano around 1005 AD. This age is older than previously assumed and suggests an emplacement only shortly after the Jala eruption. As this lava flow is considered to be the youngest one of seven post-Jala lava flows, the age also defines a period of inactivity of Ceboruco volcano of about 730-860 yr before the historic 1870 eruption. Future volcanic hazard analysis will have to take into account this result. Our work also shows that multiple sampling of single lava flows is important to obtain a reliable mean direction. Sampling sites have to be carefully selected so that they represent un-tilted parts of the flows. We interpret this to be the case for the Ceboruco lava flow, while three of the six sites of the 1870 lava flow may have been partly or completely affected by movements after thermoremanent magnetization acquisition. Unfortunately, no better sites were found for this flow.