A NEW METHOD FOR EVALUATING SHORTENING UNCERTAINTY WITH EXAMPLES FROM THE CENTRAL ANDES

Comparisons of retroarc foreland shortening in North and South America can only be meaningful if realistic assessments of uncertainty are included. Geologists frequently estimate shortening via line length balanced cross sections and state that theirs is a minimum estimate based on the inclusion of eroded hanging wall cutoffs. However, meaningful, model independent estimates would require manual construction of 100s of alternative, viable and admissible cross sections for Monte Carlo simulation, an impractical task. We propose a new method to estimate uncertainty in line length balanced sections using formal error propagation in area balancing based on complex polygons that envelop the region of interest in the deformed state and on stratigraphic thicknesses in the undeformed state. This method has the advantage that it is largely independent of internal structural interpretations, thus providing a true estimate of shortening uncertainty for a region rather than a model-dependent estimate that only includes eroded hanging wall cutoffs. We illustrate our method using an unusually well constrained part of the Subandean belt in northwestern Argentina. Shortening estimates from line length balanced sections for the Subandean belt range from 50-55 km. Using area balancing, we calculate 35 ± 20 km of shortening, not including displacement on the frontal or trailing thrusts, across the same portion of the Subandean belt. While area balancing calculates smaller shortening values, the line length balanced estimates fall within the upper bounds. Our approach highlights the true sources of uncertainty in balanced sections, in order of decreasing importance: initial stratigraphic thicknesses, depth to decollement, eroded hanging wall cutoffs. Only the last of these sources is included in traditional “minimum estimates” from line length balancing. Thus, most sections have considerably larger error than reported in the literature.