Balanced cross sections were first developed and widely deployed by the oil and gas industry to improve the success of exploration efforts in mountain belts such as the Canadian Rocky Mountains. Despite five decades of widespread use in both academia and industry, no formal method for assessing the uncertainty in these sections has ever been proposed. In this project, the investigators will remedy this shortcoming by developing the first generally applicable method to propagate errors through the calculation that leads to horizontal shortening estimates from balanced sections. Without these uncertainty estimates, it is impossible to predict the reliability, as suggested by either academic or industry practitioners, of balanced section models.
The investigators will develop formal error propagation using an area balance approach, which has the advantages that it encompasses all possible structural models, includes seldom incorporated factors such as stratigraphic thickness variation and shortening mechanisms below the resolution of the section, and can be achieved analytically, obviating the need to draw a large number of sections in the same region. Testing of the method on thrust belts in the Canadian Rockies and Taiwan will allow this research group to explore the effect of stratigraphic variations as well as kinematic styles on the degree of uncertainty in these iconic balanced cross sections.
Balanced cross sections are a fundamental tool of the structural geologist and are used extensively in oil and gas exploration, paleogeographic reconstructions, and geodynamic modeling. The cross sections are models that are fit to incomplete data and they have been used widely for oil and gas exploration in and around mountain belts for more than fifty years. Until our work, there was no formal theory for assessing the goodness of fit of the cross section or the statistical range of acceptable horizontal shortening (or extension) amounts. In this project, we have developed an approach based on area balancing that allows us to specify the input errors (uncertainties) and propagate those errors through the calculation to yield a rigorous estimate of uncertainties in shortening magnitude. Because error propagation allows us to separate out the different components contributing to the total error, we can show that most the the error does not arise from the source that most structural geologists think it does (eroded hanging wall cutoffs) but instead is due to uncertainties in the shape and thickness of the original wedge of sedimentary rocks in which the thrust belt formed. This result arises from the fact that even excellent geologic maps display significant variation in map thickness of stratigraphic units (25-35% of the mean value at the 1 sigma uncertainty level). We have not only developed the algorithms and the approach to solving this problem, but we have also written and made freely available software to enable other geologists to apply the methods in user-friendly packages. Thus, the project has provided new algorithms to the structural geology community, demonstrated that the best controlled cross sections have shortening uncertainties of 15-20% of the stated value, identified the most important sources of the uncertainty, and provided easy to use free software tools for other geologists. This project has facilitated the training of a female post-doctoral associate and initiated a discussion on a structural geology list serve regarding how best to use the new software in undergraduate teaching.
