GRADIENTS IN MINERALOGY AND ELEMENTAL COMPOSITION DEVELOP DURING CHEMICAL WEATHERING AND PHYSICAL EROSION

The increasing surface area that develops as bedrock transforms to regolith provides interfaces for extraction of nutrients and stabilization of organisms in the Critical Zone. Characteristically, regolith formation establishes mineral and elemental composition gradients that record the effects of weathering integrated over time. However, few quantitative models exist to interpret these mineralogical and elemental profiles because conceptual models to describe how bedrock transforms to regolith are generally lacking. Simulations of a four-mineral bedrock system experiencing reaction and diffusion without erosion showed that weathering leads to development of two reaction fronts. The reaction fronts define two gradients: a gradient in Fe mineralogy within an intermediate zone lying between intact bedrock and saprolite, and a gradient in silicate mineralogy in the saprolite zone. The thickness of these two zones increases with time.

Interestingly, however, when erosion is introduced at the surface at a constant rate, the width of these zones and the concentration gradients that are manifested within them are stabilized. In this case, the stationary fronts of mineral alteration move with constant and equal velocities, forming two zones of constant thickness. The thickness of the zones have been described analytically for two limiting regimes (local equilibrium and kinetic) and solutions for transient regimes have been estimated. Thus, it has been found analytically for the first time that a diffusion-limited weathering advance rate can be modeled by a rate that is constant in time when erosion occurs at a constant rate. Furthermore, it has been found that the complex multicomponent system (four-mineral model) can be split into two independent one-component reactions. Importantly, this work demonstrates that steady state mineralogical and chemical gradients can form and can be interpreted with relatively simple models.