Generate code machinery to explore the influence of parameter & observational choices on `DIVAnd` results

[Thomas-Moore-Creative]

Task is to generate code machinery to explore the influence of the following on the results of DIVAnd analyses:

• choice of correlation length scales (Lx,Ly,Lt)
• choice of epsilon2 : the noise-to-signal ratio in some cases
• choice of different types or subsets of ocean observations

Here is an example of the types of plots wanted from this code:

[Thomas-Moore-Creative]

Todo:

• [ ] start simple with single timestep (2018) and single vertical slice (2D)

working on

• https://github.com/CARSv2/cars-v2/blob/thomas-sandbox/notebooks/parameter_sandbox.ipynb based on
• https://github.com/CARSv2/cars-v2/blob/main/notebooks/Create_BG_Temperature_Field_Synthetic.ipynb which has header DIVAnd Create 2d Slice

[Thomas-Moore-Creative]

Hey @ChrisC28 - just FYI In your https://github.com/CARSv2/cars-v2/blob/main/notebooks/Create_BG_Temperature_Field_Synthetic.ipynb which I'm using I'm my code above the following was commented out as markdown but I assumed it's needed (which it appears to be)

bathname = "../src/data/TMoore-test-input/gebco_30sec_16.nc"
bathisglobal = true

dx = 0.5 # longitude resolution in degrees
dy = 0.5 # latitude resolution in degrees

lonr = 90.:dx:180.; # the range of longitudes (start:step:end)
latr = -50.:dy:0.; # the range of latitudes (start:step:end)

bx,by,b = extract_bath(bathname,true,lonr,latr)

[Thomas-Moore-Creative]

note the following from DIVAnd workshops

Looking at effets of changing analysis parameters https://github.com/gher-uliege/Diva-Workshops/blob/master/notebooks/3-Analysis/11-L-and-epsilon-effect.ipynb

Note how residuals change (decrease if you decrease epsilon ⚠️⚠️ low residuals are not necessarily a good sign as the analysis used the data points to which you compare your analysis.

⚠️ Important take-home message! Remember Analyses are not very sensitive to changes in L or ϵ if (in 2D) L * ϵ remains constant and data coverage is reasonable.

[Thomas-Moore-Creative]

Getting basic loop & plot machinery going - https://github.com/CARSv2/cars-v2/commit/4dd70cc5ddd4c033021aeb56a1b8b6866a3afc35

[Thomas-Moore-Creative]

this fed into #11 ( now closed ). @ChrisC28 has moved Julia \ DIVA development further towards completion.