Coastal Downscaling Experiments: Can CESM Fields Successfully Force Regional Coastal Ocean Simulations with Strong Freshwater Forcing?

Tuesday, May 13, 2014 - 07:00
Add to Calendar

The coastal ocean accounts for about half of the global fish harvest, but is poorly resolved in global climate models (a one-degree grid barely sees the continental shelf). Moreover, coastal ocean circulation is strongly modified by river freshwater sources, often coming from estuarine systems that are completely unresolved in the coarse grid. River freshwater input in CESM is added in a practical, but ad hoc, way, by imposing a surface salinity sink over a region of the ocean approximating the plume area of a given river. Here we present results from a series of model experiments using a high-resolution (1.5 km) ROMS model of the NE Pacific, including the Columbia River and the inland waters of Puget Sound. The base model does multi-year hindcasts using the best available sources of atmospheric (MM5/WRF), ocean (NCOM), river (USGS), and tidal forcing. It has been heavily validated against observations of all sorts, and performs very well, so it is an ideal test bed for downscaling experiments. The model framework also does biogeochemistry, including oxygen, and carbon chemistry is being added to make forecasts of Ocean Acidification. This high-resolution ROMS model is systematically run in downscaling experiments for the year 2005 with combinations of CESM forcing (CAM, POP, and rivers) swapped in. Skill is calculated using observations. It is found that the runs with CESM forcing generally retain much of the skill of the base model. A compact metric of response to freshwater forcing is developed, which is the mechanical energy required to destratify a shallow coastal volume. This, along with the average temperature and salinity of the volume, are used to characterize and compare runs, including the original CESM-POP fields. Finally the model is run with projected CESM simulation forcing at the end of 21st century based on a set of RCP scenarios, and the compact metrics are used to quantify differences from 2005.

Presentation file(s):