Linear compositional modeling and the statistical significance of alteration intensity factors, Clark segment, San Jacinto fault zone, Horse Canyon, California: Implications for the illite/smectite to illite transition and the development of fault zones at shallow crustal levels
Wednesday, May 8th, 2013
The NE block of the Clark segment of the San Jacinto fault zone within Horse Canyon, NE of Anza, southern California displays a well-defined damage zone, transition zone, and fault core. Previous work by N. Morten and colleagues (2012) showed that through fragmentation and slip grain and fragment size progressively diminishes from the outer damage zone inward toward the fault core where the finest grain size material is found. In addition, the proportion of plagioclase declines while the proportion of quartz increases towards the fault core. When chemical data are plotted on a A-CN-K ternary diagram a linear trend extending from wall rock compositions toward the compositional field of illite is evident. In addition, N. Morten and colleagues showed through evaluation of the <4 micron fraction, that the illite/smectite to illite transition occurred at or near the transition zone/fault core boundary. However, because of the coarseness of <4 micron fraction, XRD data included peaks from quartz, plagioclase, K-feldspar, and amphibole along with those derived from the clay minerals, all making interpretations difficult and at times problematical. We therefore undertook a clay mineralogy study of the <2 micron fraction, and revaluated the chemical alteration trends first recognized by Morten and colleagues utilizing the linear compositional modeling techniques promulgated by H. von Eynatten and colleagues in 2003. From the latter work we determined the translational invariant alteration intensity factors for the damage zone, transitional zone, and fault core of the NE block.
The results of our work shows that the <2 micron fraction provides a more concise and cleaner interpretation of the illite/smectite to illite transition within the NE block. In addition, our modeling using non-central component analysis shows that A-CN-K data spread about a compositional linear trend with PC1 explaining 99.7% of the simplicial variability. Through orthogonal projection we calculated the alteration intensity factors for each sample analyzed from the various architectural components of the fault zone. In order to assess whether or not the means of the alteration intensity factors were statistically different at the 95% confidence level, we utilized the one-way ANOVA routine in SPSS. The significance level of the omnibus results was .0001; hence, there is not sufficient statistical evidence to reject the null hypothesis that at least one of the means was different than the other two. We then utilized the post hoc routines in SPSS which revealed that the mean of the alteration intensity factors for the fault core are different from the means of those obtained from the transition and damage zones at the 95% confidence level. In contrast, the means of the transition and damage zones are not different at the 95% confidence level. The above results suggest that for fault zones derived from tonalitic wall rocks at depths of ~0.5 km, the illite/smectite to illite transition will occur when alteration intensity factors exceed 0.20, the average intensity factor calculated for the transition zone. We speculate that under such conditions during repeated rupturing events fluids with elevated temperatures (≥ ~125oC) are flushed through the fault core. In short, over time, the combination of shearing, fragmentation, and relatively elevated temperatures eventually overcomes the kinetic barrier for the illite/smectite to illite transition. Such settings and processes are unique to fault zones, and as a result, fault zones represent an underappreciated setting for the development of illite from illite/smectite.