New article on stress inversions

Diversity is a virtue, even for stress inversions.

That is the primary finding of a new paper by me, Eric Beaucé, and Orlando Teran in a Geological Society of London Special Publication on stress. Using synthetic earthquake focal mechanisms based on real earthquakes that have occurred recently in the Permian Basin of West Texas, we show that low nodal plane orientation diversity can result in strongly inaccurate stress inversion results. The figure shows how the results of stress inversions (and their uncertainty ranges) differ from the input (true) stress values for several groups of focal mechanisms with different nodal plane orientation diversity. As is evident in the figure, where nodal plane orientation is low, bootstrap sampling can underestimate the true uncertainties, meaning that users may not realize that their results are compromised. This problem is compounded with few focal mechanisms or noisy (low quality) focal mechanisms.

In addition to the results, we provide some tools, including use of the resolution and posterior covariance operators, that allow users to judge the quality and uniqueness of their own stress inversions. We hope that this study and the accompanying visual and software tools will make it easier for users to obtain high-quality stress inversion results.