docs: add notebooks for modpath7 examples 3-4

[wpbonelli]

Add notebooks for problems 3 and 4 from the examples bundled with the MODPATH 7 distribution

• example 3
• example 4

Both examples are reproduced programmatically.

Could later be updated to compare MODPATH and PRT (when released)

[codecov]

Codecov Report

Merging #1687 (5724880) into develop (7aa9ecf) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           develop   #1687   +/-   ##
=======================================
  Coverage     71.9%   71.9%           
=======================================
  Files          253     253           
  Lines        55987   55987           
=======================================
  Hits         40270   40270           
  Misses       15717   15717           

[jdhughes-dev]

Using edgecolor or facecolor will eliminte the warning in flopy_modpath7_structured_transient.ipynb which is causing the warning that is showing your file absolute path. Did we ever find a way to scrub this type of data?

If you reverse the order of your polygons in flopy_modpath7_unstructured_transient.ipynb will this eliminate the GeoJSON wayning?

[wpbonelli]

Thanks @jdhughes-usgs, fixed warnings.

Absolute paths were cleaned up in #1711, but it looks like paths aren't always computed correctly now (e.g. executable paths in notebooks here). Opened #1728 to address

[wpbonelli]

Fixed executable paths here as per approach proposed in 1728

[christianlangevin]

Looks great, @w-bonelli! Few minor things for your consideration.

• "stress_period_data={0: dd, 1: dd, 2: dd}" If the drains are the same for every stress period, you only need to specify them for the first one and MODFLOW will automatically keep applying them.
• The figure that uses the add_patches is a little confusing. The legend has color fill symbols, but not polygon symbols. And maybe features are overlying one another, but the I don't see all the legend colors on the plot. Maybe a combination of symbols and filled polygons would be best?
• We often shrink the colorbar using shrink=0.5
• The unstructured notebook looks great.

[wpbonelli]

Thanks @langevin-usgs, just added revisions:

• update drain stress period usage
• shrink color bar by 0.5
• fix legend/patch color consistency
• lower alpha for heads on the last plot so patches can be seen more clearly
• use plot_bc("WEL", plotAll=True) instead of custom patch to show wells
• remove lingering username from output

[wpbonelli]

Fixed pathline colors in existing notebooks to address #1732, also updated to suppress/capture model logs and avoid print statements showing executable path (per #1728)