Patch SNOWT_AVG and ACSNOM

[GreyREvenson]

Purpose

To fix handling of SNOWT_AVG and ACSNOM variables that were added to the gridded branch in PR #86.

Changes

• SNOWT_AVG was being set to huge(1.) when there was no snow/ice. However, this PR changes the way SNOWT_AVG calculated such that SNOWT_AVG will be -999999.0 when there is no snow/ice. See PR Add support for NWM 3.0 baseline variables #91 and this comment specifically for further explanation.

• This PR adds SNOWT_AVG to energy_type%InitDefault, energy_type%TransferIn, and energy_type%TransferOut because SNOWT_AVG should have been present in these subroutines but was not. This problem surfaced when I reexamined the unit test output for SNOWT_AVG (after changing our means of calculating SNOWT_AVG as described in the first bullet point above) but did not see expected values for the variable in the unit test output.

  • Note that PR Add support for NWM 3.0 baseline output variables #86 included SNOWT_AVG in the transfer in/out subroutines but this change was mistakenly undone by PR Make 'transfer-in' and 'transfer-out' subroutines type-bound to noahowp_type sub-types #89

• This PR adds ACSNOM to water_type%InitDefault, water_type%TransferIn, and water_type%TransferOut because ACSNOM should have been present in these subroutines but was not. This problem surfaced when I reexamined the unit test output (after changing our means of calculating SNOWT_AVG as described in the first bullet point above) but did not see expected values for the variable in the unit test output.

  • Note that PR Add support for NWM 3.0 baseline output variables #86 included ACSNOM in the transfer in/out subroutines but this change was mistakenly undone by PR Make 'transfer-in' and 'transfer-out' subroutines type-bound to noahowp_type sub-types #89

Testing

• Re-executed unit test and got this output.

[nmizukami]

Hi Grey,

This is small PR and this should be good to go. I put one comment (something to think about for future).

[nmizukami]

maybe like this? realMissing looks like parameter (no value change)

REAL, parameter :: realMissing=-999999.0

and then remove L37 (realMissing = -999999.0). this is more like parameter so not in the data initialization category.

[nmizukami]

Also, this would be personal choice, but parameter variable could be stored in a separate module so they can be shared from there. if realMissing is used for somewhere and you have many similar global parameters, this would be way to go.

[GreyREvenson]

Good point. I went ahead and set the value of realMissing when it is declared, as you suggested.

I also agree that we should implement a constants module at some point. I know it's an issue (i.e., Issue #5) -- I just need to get with Keith regarding prioritization of tasks.

[GreyREvenson]

Thank you for having a look, @nmizukami. Appreciate it!

Apologies for taking so long to get back to you. I was distracted by AGU.