Influence of measurement errors on stress estimated from single-phase fault/slip data

The influence of measurement errors on estimated stress is investigated in reduced sigma space on the basis of numerous sets of homogeneous fault/slip data generated under different extensional stress and then modified by varying measurement errors for the striation pitch. For a given measurement error, error propagation to a datum vector depends upon the controlling stress and fault orientation with respect to the controlling stress. For a given stress ratio, both the dispersion of datum vectors (i.e. root mean square cosine) and the misfit between estimated and controlling stresses (i.e. similarity and stress difference) increase with the increase in measurement error. The two measures appear generally smaller at small measurement error than at large error and have a strong tendency to increase with stress ratio for a given measurement error. The measurement error has an apparent effect in estimated stress only if it is greater than 10°, implying a limited effect on homogeneous fault slip data that is carefully measured at outcrop, in which a smaller error can be guaranteed in a majority of cases.