Comparison to CO2SYS

[oscarbranson]

After Orr et al (2015).

[tompc35]

@oscarbranson Thanks for putting the effort into this project. It looks great. I could potentially help with the comparison to CO2SYS (or one of the other issues on the TODO list). I am a physical oceanographer, and have recently transferred most of my work from Matlab to Python.

[oscarbranson]

Thanks @tompc35 - that would be really useful. I've actually made a start on this, but would welcome help. I was thinking of comparing cbsyst to GLODAP/WOA data, after Orr er al (2015).

I'd like the comparison to be self-contained, so any user can run it on their own system - i.e. get the data, run it through CO2SYS.m, and generate a 'comparison report' containing plots comparing cbsyst to CO2SYS.m (e.g. scripts making the GLODAP plots in the readme). This could be wrapped in two scripts, and the workflow could be something like:

1. run get_GLODAP_WOA_data.py (this could generate some matlab code as output?)
2. user runs data through CO2SYS, saves output in same directory (with comparable constants)
3. run plot_CO2SYS_comparison.py

So far, I've written get_GLODAP_WOA_data.py, which scrapes and combines the GLODAPv1/WOA data, and I've run this through CO2SYS.m so have a comparison dataset offline. plot_CO2SYS_comparison.py is where work is needed.

All that said, it might be worth holding off a couple of weeks until #13 and #8 are closed, which will add the final features to to match the full capabilities of CO2SYS.m. So, I think a roadmap to getting this is done is something like:

• [ ] Implement separate input/output TPS conditions (#8)
• [ ] Implement Revelle factor reporting (#13)
• [ ] Write plot_CO2SYS_comparison.py
• [ ] If Alternative Constants (#14) are added, modify comparison scripts to allow for this.

I've created a branch to work on this called co2sys_compare. Feel free to contribute! Happy to discuss further on here.