Dynamic downscaling boundary condition generation
Scientific Motivation
In some regions, small scale processes that are not captured in global models dominate trends in biogeochemistry. For example, ocean acidification is the predictable global response to rising atmospheric carbon dioxide concentrations from anthropogenic emissions. The open ocean pH follows the trend in atmospheric carbon dioxide concentration, declining at a relatively stable rate of 0.0019/year. However, in the coastal ocean, local processes can modulate or mask the trend. Constraining the carbon budget in the coastal ocean is especially worthwhile because it is home to several of the most vulnerable species to ocean acidification. A similar argument can be made for oxygen decline. Therefore, regional high-resolution models are necessary to forecast and project ocean conditions nearshore. These regional models require boundary conditions from global models. This hack will facilitate boundary condition generation from CMIP6 models.
Proposed Hacking
Build a tool that generates boundary conditions for a regional model. Users input the coordinates of the spatial boundaries of their regional model, as well as the time period of interest and temporal resolution. The tool will extract subsurface fields along boundaries. Optionally, this tool can interpolate to the model grid.
I am familiar with the ROMS regional modeling tool, so I was thinking the output could be in ROMS format, but I am open to other ideas.
This tool can make it easier to create a regional model ensemble forced by boundary conditions from different CMIP6 models.
Anticipated Data Needs
Subsurface monthly ocean:
physical: sea surface height, salinity, temperature, velocity,
biogeochemical: DIC, TA, O2, NO3
Anticipated Software Tools
xarray, intake_esm, looking for ideas
Desired Collaborators
I am interested in collaborating with anyone with an interest in biogeochemistry and/or regional modeling. I’m inexperienced with global earth system models so any knowledge of them is also appreciated!