Statistical data analysis of the CCDB (Collapse Caldera Database): Insights on the formation of caldera systems

Collapse calderas are one of the most important volcanic structures because of their hazard implications, but also because of their high energy potential and their association with mineral deposits of high economic interest. The worldwide Collapse Caldera Database (CCDB) [22], ( http://www.GVB-csic.es/CCDB.htm), currently formed by 473 calderas and 28 variables, updates the current field based knowledge on calderas, merging together the existing databases, complementing them with new examples found in the bibliography and leaving it open for the incorporation of new data from future studies. The area of the caldera is a relevant indicator of the dynamics of the caldera system, as it is directly related to the size of the magma chamber and other variables of the volcanic systems such as crustal type, magma composition, and tectonic setting. Despite this being a generally assumed geological belief, it has never been numerically studied. In this paper we have conducted a statistical analysis using ANOVA (Analysis of Variance) on a selected sample from the CCDB containing the relevant variables, in order to determine if the area of the collapse caldera depends on a group of geological characteristics or is it due to random variability alone, and if so, what is the relationship. The results identify three groups of collapse calderas in different geodynamic environments according to Crustal Type, Rock Suite and Plate Tectonic Setting, which we name GE1, GE2 and GE3. These groups have significantly different Caldera Area, suggesting how the dynamics of the volcanic system hosting the collapse caldera are intrinsically related to the geodynamic environment where it develops.