Youqing Yang

Crustal shortening in the Andes: Why do GPS rates differ from geological rates?

GPS data indicate 30–40 mm yr−1 present-day crustal shortening across the Andes, whereas geological evidence shows crustal shortening concentrated in the sub-Andean thrust belt at a much lower rate (<15 mm yr−1). We reconcile the discrepancy between the geodetic and the geological crustal shortening using geodynamic modeling that includes timescale-dependent crustal deformation. The GPS velocities reflect the instantaneous deformation in the Andes that includes both permanent deformation and elastic deformation that will be recovered during future earthquakes, whereas the lower geological rates reflect only the permanent deformation. The three-dimensional viscoelastic model predicts nearly uniform short-term velocity gradients across the Andes, similar to the GPS results, and concentrated long-term crustal shortening in the sub-Andean thrust zone, consistent with geological observations.

A 3-D geodynamic model of lateral crustal flow during Andean mountain building

Although the Andes are believed to have resulted mainly from crustal shortening, the shortening history remains debated and appears to require lateral (along-strike) crustal flow. Three-dimensional viscous flow modeling shows that, within geological uncertainties, the Andes may have been produced by either Neogene shortening alone or with significant pre- Neogene shortening. These scenarios require major along-strike crustal flow and predict significantly different histories of uplift and crustal motion.

Algorithms for Optimizing Rheology and Loading Forces in Finite Element Models of Lithospheric Deformation

Lithospheric deformation results from dynamic interplay of tectonic driving forces (loading) and lithospheric properties (rheology and structure). Unlike engineering problems, in lithospheric dynamics both the loading conditions and lithospheric properties are often hard to constrain, forcing many computer models to oversimplification. These simplified models usually cannot take the full advantage of the fast growing observational constraints. In this article, we present algorithms that help to seek optimal loading conditions and rheological parameters in models of lithospheric deformation. In particular, we use genetic algorithms to iterate for the optimal rheological structure, and a regression algorithm for optimizing tectonic loading. We illustrate these algorithms in two models: a plate flexure and a viscous three-dimensional lithospheric deformation. In both cases these algorithms utilize the observational constraints to obtain the optimal driving forces and lithospheric rheology. The results significantly improve over those derived from traditional approaches.