We are proposing a dual-strategy approach for investigating the role of coastal and estuary processes in the cycling of freshwater through the Earth system and improving their representation in the Community Earth SystemModel (CESM). The first is a dynamical downscaling approach through nested regional modeling. We will use this capability to downscale decadal timeslice climate hindcasts and predictions for the coastal and inland waters of the Northeast Pacific for the early and late 21st century to advance our understanding of the response of coastal environments to global modes of climate variability and climate change, with particular focus on the relative importance of local river runoff, local surface forcing, and large-scale ocean circulation. The second approach will be to develop computationally efficient, but physically based, parameterizations of estuary and continental shelf mixing processes that can be added as new modules to the standard (non-nested) configurations of CESM, allowing for straightforward application to freshwater inputs throughout the world. We will use the new parameterizations in 20th and 21st century climate simulations to advance our understanding of the magnitude, range, and timing of upscale impacts of freshwater inputs to the coastal ocean on the Earth system with particular emphasis on upper ocean stratification, water mass formation, and the response of the global ocean to a changing hydrologic cycle in a warming climate.The proposed effort will advance the state-of-the-science in Earth system modeling in several ways. We will provide, for the first time, a representation of estuaries and shelf processes in global models, thereby reducing biases in coastal stratification and provide a key link in the global water cycles that has been ignored to this point. We will provide a coupled, dynamical downscaling capability for coastal environments – an element of the Earth system that is critical to both natural ecosystems and sustainability of the human population. The analysis of our experiments will provide insights into the mechanisms of decadal variability for these environments, and how coastal processes influence global modes of climate variability and climate change. The parameterizations and nested-modeling infrastructure to be developed under the proposed effort will be incorporated into the publically available CESM software distribution, with all necessary documentation and input data. This represents a significant enhancement in the capabilities of this widely used modeling system, opening the opportunity for a broad range of new research in areas such as climate impacts on coastal ecosystems.