add surface flux tutorial to examples

[wilsonduan10]

Purpose

This PR adds another example to the EKP docs which details the use of EKP to recover surface flux parameters in a perfect model example.

Content

This PR adds a few files in the examples/SurfaceFluxExample folder:

kappa_calibration.jl

This file implements EKI to recover the von-karman constant, which is used in many MOST equations. It pulls data from the John Hopkins Tubulence Channel Flow, which is used by ensemble members to predict friction velocity u*. The difference in the predicted u* to the observed u* given by the dataset is used to calibrate the von-karman constant. This file is rendered to the documentation through Literate and Documenter along with a few other example Julia files.

This file differs from the file given in the SURF_2023 folder in a few ways:

• formatting: moved all imports to the top
• only downloaded the profiles data since the code did not use the standard deviation data
  • however, we could use the standard deviation data to add noise to the input u
• made code in the physical_model more concise
  • the physical model attempts to override the stability function constants with the parameter von karman constant, but there is nothing to override
  • the physical model extracts the von karman constant from the toml only to override it in a following line, so it is unnecessary to extract the von karman constant in the first place
• changed the Gamma noise constant to scale by 0.0005 instead of 0.1 because 0.1 is too much noise; the observable u_star_obs is only ~0.04, so adding noise on the order of magnitude of 10^-1 is too much noise. This noise reduced the model's effectiveness.
• In addition to plotting the truth observed u_star, I plotted y as well as the output of the model given the truthful parameters
• I changed the second plot to a histogram to better fit the data given
• I removed the transform_init_prior_to_constrained function because there is already a method provided by EKP that gets constrained parameters

setup_parameter_set.jl

This file is a helper file used by kappa_calibration.jl to define functions that are helpful for setting parameters necessary for the surface_conditions function from the SurfaceFluxes package.

alt_kappa_calibration.jl

This file is an alternate method of recovering the von-karman constant. Instead of using surface_conditions as the forward model, this example simply uses the wind profile function. The observable used is the mean wind given by the same dataset as kappa_calibration.jl. Although this file is not rendered to the documentation, it can be used by new developers to better understand the EKI pipeline.

---

• I have read and checked the items on the review checklist.