Growth of the Phoenix metropolitan area led to exposures of the internal bedrock structure of surrounding semi-arid mountain ranges as housing platforms or road cuts. Such exposures in the Sonoran and Mojave deserts reveal the presence of sedimentary calcium carbonate infilling the pre-existing fracture matrix of the bedrock. Field surveys of bedrock fractures filled with carbonate (BFFC) reveal an average of 0.079 +/- 0.024 mT C/m2 stored in the upper 2 m of analyzed bedrock exposures. Back-scattered electron microscopy images indicate the presence of carbonate at the micron scale, not included in this estimation. Analysis of the spatial extent of bedrock landforms in arid and semi-arid regions worldwide suggests that ~1485 GtC could potentially be stored in the upper 2 m horizon of BFFCs. Radiocarbon dating obtained at one of the sites indicates it is likely that some of the carbonate was flushed into the bedrock system during glacial wet pulses, and is stored on Pleistocene timescales or longer. Strontium isotope analysis at the same site suggest the potential for a substantial cation contribution from weathering of the local bedrock, indicating the potential exists for sequestration of atmospheric carbon in BFFCs. Rates of carbon release from BFFCs are tied to rates of erosion of bedrock ranges in desert climates.

Fluctuating flow releases on regulated rivers destabilize downstream riverbanks, causing unintended, unnatural, and uncontrolled geomorphologic changes. These flow releases, usually a result of upstream hydroelectric dam operations, create manmade tidal effects that cause significant environmental damage; harm fish, vegetation, mammal, and avian habitats; and destroy riverbank camping and boating areas. This work focuses on rivers regulated by hydroelectric dams and have banks formed by sediment processes. For these systems, bank failures can be reduced, but not eliminated, by modifying flow release schedules. Unfortunately, comprehensive mitigation can only be accomplished with expensive rebuilding floods which release trapped sediment back into the river. The contribution of this research is to optimize weekly hydroelectric dam releases to minimize the cost of annually mitigating downstream bank failures. Physical process modeling of dynamic seepage effects is achieved through a new analytical unsaturated porewater response model that allows arbitrary periodic stage loading by Fourier series. This model is incorporated into a derived bank failure risk model that utilizes stochastic parameters identified through a meta-analysis of more than 150 documented slope failures. The risk model is then expanded to the river reach level by a Monte Carlos simulation and nonlinear regression of measured attenuation effects. Finally, the comprehensive risk model is subjected to a simulated annealing (SA) optimization scheme that accounts for physical, environmental, mechanical, operations, and flow constraints. The complete risk model is used to optimize the weekly flow release schedule of the Glen Canyon Dam, which regulates flow in the Colorado River within the Grand Canyon. A solution was obtained that reduces downstream failure risk, allows annual rebuilding floods, and predicts a hydroelectric revenue increase of more than 2%.