add checks on snow states after application of LDAS increments; StieglitzSnow cleanup

[gmao-rreichle]

Summary of changes:

• Adds checks on snow states after application of LDAS increments from MODIS snow cover assimilation in coupled land-atm system (catch_incr.F90).

• Adds FICE[n] exports from Catchment (supersedes additional snow module exports from Catchment #814)

• 0-diff cleanup and documentation of StieglitzSnow.F90:

  • adds missing "implicit none" to StieglitzSnow module
  • improves handling and naming of some public/private constants (CATCH_SNOW_*)
  • changed interface and functionality of StieglitzSnow_relayer() for use in LDAS
  • revises interface of snowrt() to group arguments with matching intent; add missing intent statement
  • fixes indentation (best to hide white-space differences when viewing changes!)
  • removes obsolete (commented-out) lines of code and constants
  • use StieglitzSnow_calc_asnow() instead of writing out equation each time
  • adds "module procedure" interface for StieglitzSnow_calc_asnow() to facilitate scalar and vector inputs and outputs
  • adds "module procedure" interface for StieglitzSnow_calc_tpsnow() for scalar/vector inputs [but maintains different if condition thresholds used in the original scalar (get_tf0d) and vector (get_tf_nd) subroutines for now
  • renames some public StieglitzSnow subroutines for clarity
  • cleanup of module use statements related to StieglitzSnow
  • adds documentation
  • removes commented-out and obsolete copy of StieglitzSnow subroutines from GEOS_LandIceGridComp.F90 to avoid confusion
  • replaces hard-wired number of snow layers ("3") with parameter in GEOS_Catch[*]GridComp.F90

• Fixes file permissions of SurfParams.F90 and GEOS_SurfaceGridComp.rc.

• Removes obsolete variables and use statements (catchment[CN].F90, GEOS_LandIceGridComp.F90, lsm_routines.F90)

• Removes suspicious catchcn_calc_*() subroutines in catchmentCN.F90

---

Testing to date:

• tested successfully for GEOSldas on Oct 6 (after 667edae) by @gmao-rreichle (requires eponymous GEOSldas add checks on snow states after application of LDAS increments GEOSldas#676); tests with aggressive compilation have round-off differences

• retested successfully for GEOSldas Intel-only on Oct 15 (after 48ff3cf) by @gmao-rreichle

• fully tested by @biljanaorescanin on 16/17 Oct 2023

  • All AGCM tests passed. (AMIP, REPLAY, No Inc Replay and Regression)
  • GEOSldas tests passed except for "aggressive" tests, which failed the comparison with roundoff differences, as expected.

---

@rdkoster @biljanaorescanin @weiyuan-jiang @lcandre2 @saraqzhang

[gmao-rreichle]

Converting back to draft so I can finish the non-essential cleanup, now that we're not furloughed.

[biljanaorescanin]

@sdrabenh this PR is ready for your review.
I did standard testing on it (amip, replay and zero increment replay) and all tests passed.