Fix backscatter divide by zero

[adcroft]

A bug fix and an associated cleanup:

• Commit f7fd4e4 addresses a divide by zero encountered by @ElizabethYankovsky in the NeverWorld2 configuration while trying to use back-scatter.
• Commit fc3e3ac cleans up the capitalization of indices to follow https://mom6.readthedocs.io/en/dev-gfdl/api/generated/pages/Horizontal_Indexing.html#soft-convention-for-loop-variables

One question for @Hallberg-NOAA : could we ever encounter a divide by zero indicated by comment at top of fc3e3ac ?

[codecov]

Codecov Report

Merging #62 (d3e8970) into dev/gfdl (9f0018f) will not change coverage.
The diff coverage is 0.00%.

@@            Coverage Diff            @@
##           dev/gfdl      #62   +/-   ##
=========================================
  Coverage     28.94%   28.94%           
=========================================
  Files           242      242           
  Lines         71553    71553           
=========================================
  Hits          20714    20714           
  Misses        50839    50839           

Impacted Files	Coverage Δ
src/parameterizations/lateral/MOM_hor_visc.F90	43.63% <0.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 9f0018f...d3e8970. Read the comment docs.

[herrwang0]

Not sure if this is relevant, but I think h_u and h_v could be zero (rather than an Angstrom) over the masked-out region when USE_LAND_MASK is True (which is the default). If so, hrat_min would be zero as well.

MOM6/src/parameterizations/lateral/MOM_hor_visc.F90

Line 616 in 6da5c9b

h_u(I,j) = 0.5 * (G%mask2dT(i,j)*h(i,j,k) + G%mask2dT(i+1,j)*h(i+1,j,k))

I never checked the actual values of h_u and h_v over the land though, just something that occurred to me when I was going through this piece of code earlier.

[marshallward]

Following on @herrwang0's comment, could this also happen here?
Nevermind, I just saw the comment in the code.

MOM6/src/parameterizations/lateral/MOM_hor_visc.F90

Lines 1000 to 1009 in 6da5c9b

	if (CS%better_bound_Kh) then
	do j=Jsq,Jeq+1 ; do i=Isq,Ieq+1
	if (Kh(i,j) >= hrat_min(i,j) * CS%Kh_Max_xx(i,j)) then
	visc_bound_rem(i,j) = 0.0
	Kh(i,j) = hrat_min(i,j) * CS%Kh_Max_xx(i,j)
	else
	visc_bound_rem(i,j) = 1.0 - Kh(i,j) / (hrat_min(i,j) * CS%Kh_Max_xx(i,j))
	endif
	enddo ; enddo
	endif

[marshallward]

Since the potential error in the other block is denoted in the comments, and since we're getting ready for a merge to main, I will probably just merge this one and we can address it in a later version.