A complex relationship exists between climate, tectonic uplift, and erosion. The Andean Plateau is an excellent site for studying these processes, their interactions, and the geologic history of Andean mountains and plateau.

Climate and Erosion Patterns
The elevation and morphology of mountain belts is a balance between tectonic uplift and surface erosion. The climate of a region is a key control on the amount, style and location of surface erosion and therefore mountain belt morphology. Because climate changes through time, we cannot assume that the processes acting today were the same throughout the development of a mountain belt. The Andes Mountains have grown and developed over the last 40Ma. During this time the regional climate has changed in direct response to Andean uplift and also in response to a changing global climate.

Our work quantifies the climatic change in South America during the growth of the Andes over the last 40Ma in order to determine the influence of this climatic change on erosion patterns.  Regional Climate Models are used to examine the impact of atmospheric CO2 level, the presence of ice sheets, the elevation of the Andes and vegetation patterns on South American climate. The results of the climate models are used as input into landscape evolution models to assess the contribution of climatic change to changes in erosional regimes.

The Andean Plateau is one of the most dramatic topographic features on Earth. Despite its impressive nature, the mechanisms and rates of its formation remain poorly understood. A range of competing geodynamic models have been proposed to explain the formation of plateaus. These models call upon processes such as delamination of a dense root, crustal shortening and thickening, under-plating, magmatic addition, ablative subduction, lower crustal flow, and a coupling between lithospheric and atmospheric processes through erosion. To distinguish between these dynamic models, observational constraints on the evolution of the Andean Plateau are needed. These constraints come from an understanding of: (1) the kinematic (deformation) and erosional history associated with the plateau and marginal thrust belts, (2) the present day crustal and lithospheric structure, and (3) the paleoelevation history of the plateau.

A) Modern topography of western South America. B) Contending geodynamic models for plateau surface uplift. Light gray represents over thickened low-density crust. Red dashed lines indicate regions where higher-density lithospheric mantle material (dark gray) is removed, thereby resulting in surface uplift. (Ehlers and Poulsen, 2009)

Of these, our understanding of Andean paleoelevation is arguably the most uncertain. The paleoelevation history can be used to make inferences about the rate and timing of mountain building, and ultimately the geodynamic processes that govern mountain building. Though several methods (e.g. using foliar physiognomy, stomatal density in fossil leaves, stable isotopes in soil minerals, vesicularity of basalitic flows) have been used to infer paleoelevation, stable isotope paleo-altimetry is considered the most direct method due to the strong correlation between air temperature and precipitation δ18O. This method has been used to provide detailed records of paleo-altimetry for the North American Western Cordillera, European Alps, and the Tibetan and Andean Plateaus. In the Bolivian Andes, Garzione et al. [2006] and Ghosh et al. [2006] have used this method to infer a rapid rise of the plateau to its modern elevation between ~10.3 and ~6.7 Ma, a history which may be best explained by surface uplift due to the removal of a dense lower crust and/or lithospheric mantle.

The paleoelevation history of the Andes remains controversial however, because is at odds with other lines of geological evidence. Using mass-independent fractionation anomalies (17O) in paleosols as a proxy of aridification, Rech et al. [2006] suggest that the Atacama Desert reached an altitude of >2 km between 19-13 Ma. Furthermore, thermochronometer samples tied to balanced cross sections in northern and southern Bolivia suggest that the Andean Plateau reached its modern width (but unknown elevation) by ~20 Ma and experienced relatively constant shortening rates from ~15 Ma to present. The presence of a dense mantle root could account for the low elevation of the plateau prior to ~10 Ma; however, the constant shortening rates and the initiation of Subandean deformation since ~20 Ma are seemingly inconsistent with a rapid Andean rise.

We are working to resolve this controversy, and the timing of the uplift of the Andean Plateau, by rigorously evaluating the 18O paleo-altimetry methodology and its application in the Andes. The 18O paleo-altimetry method assumes that the isotopic concentration of an air parcel is fractionated according to Rayleigh distillation, and is not significantly influenced by mixing with surrounding air masses or by changing source regions. There are several reasons to suspect that these assumptions may not be entirely valid in the Andes. First, records of modern precipitation δ18O and climate simulations of modern precipitation δ18O from the Andean Plateau do not strictly follow a Rayleigh distillation model. Second, our results using regional atmospheric circulation models suggest that the source and path of Andean vapor changed substantially in response to the rise of the Andes.

Our research has two components: 1) to improve our understanding of the modern δ18O cycling over the central Andes through both field campaigns and climate modeling, and 2) to develop paleoclimate simulations to estimate past changes in precipitation 18O. To understand the modern isotopic system, we have installed and are monitoring a network of weather and rain gauge stations across central and southern Bolivia. The data from this campaign will provide much-needed information about the source and variability of rainfall δ18O in the central Andes. We are also developing and using regional climate models to predict the modern and past δ18O climatology over the Andes. These models can help us untangle the many processes that influence rainfall δ18O.