Fix the bug: TUV photolysis is inactive if simulation starts on 1st Jan

[SeregaOsipov]

Fix the bug: photolysis is inactive if simulation starts on 1st Jan

TYPE: bug fix

KEYWORDS: TUV, distance to the sun, 1 January, Julian day

SOURCE: Sergey Osipov (KAUST)

DESCRIPTION OF CHANGES:
Problem:
TUV and FTUV fail to initialize the distance to the Sun properly if the simulation starts on 1 Jan. The multiplication factor remains uninitialized at 0, which zeros out photolysis rate calculations (see line 745 @ module_phot_tuv.F. On the next day, the distance is recalculated normally and photolysis rates are calculated properly.

Solution:
Given that the calculation of sun distance is trivial, the if check was disabled

ISSUE:

LIST OF MODIFIED FILES:
chem/module_phot_tuv.F
chem/module_ftuv_driver.F

TESTS CONDUCTED:

1. Do mods fix problem? How can that be demonstrated, and was that test conducted?
2. Are the Jenkins tests all passing?

RELEASE NOTE: Include a stand-alone message suitable for the inclusion in the minor and annual releases. A publication citation is appropriate.

[weiwangncar]

The regression test results:
Test Type | Expected | Received | Failed
= = = = = = = = = = = = = = = = = = = = = = = = = = = =
Number of Tests : 23 24
Number of Builds : 60 57
Number of Simulations : 158 150 0
Number of Comparisons : 95 86 0

Failed Simulations are: 
None
Which comparisons are not bit-for-bit: 
None

[weiwangncar]

@jordanschnell Just wondering if you could help review this PR? Thanks!

[jordanschnell]

Hi Sergey, thanks for this fix. Instead of commenting out the check such that it is called each time, what about including a check for julday = 0?

[SeregaOsipov]

Hi Jordan, good to hear from you.

Julday is zero if the simulation start date is 1st Jan. At least, this is what I saw in the DDT debugger on our supercomputer.
The value is set in line 346 in chem_driver.F: ijulian=ifix(grid%julian) and then passed to the photolysis_driver.F

An alternative is to initialize the curjulday with -1 (or -999) instead of 0, which is used as a cache for the last calculation.

Overall, the sundist calculation is very trivial and only involves trigonometric functions, meaning that the computatinal overhead is negligeble. IMHO, this is a good trade-off to avoid the hard-to-predict bugs like the 1st January case.

Best, Sergey

[jordanschnell]

OK with me. Approved.

The base branch was changed.