Refactor SAMF convection schemes for single-precision (32-bit) capability

[rhaesung]

Description of Changes:

• Refactored internal calculations, working precision, and local variables in samfdeepcnv.F90, samfshalcnv.f, progomega_calc.f90, and progsigma_calc.f90 to 32-bit (conv_wp) to reduce memory bandwidth and enhance computational performance.
• Maintained 64-bit double precision (kind_phys) for external I/O arrays and host model interfaces to ensure seamless integration and stability with the host workflow.
• Updated CCPP metadata (.meta) files and variable specifications to ensure accurate glue-code generation for flexible precision switching.

Tests Conducted:

• 64-bit Baseline Verification: PASSED. Achieved 100% Bit-for-Bit (B4B) numerical agreement against double-precision baselines across all regression tests.
• Scientific Validation: 32-bit scientific accuracy testing and global workflow evaluation are currently underway.

Dependencies:

• NOAA-EMC/ufsatm#<pr_number>
• ufs-community/ufs-weather-model/#<pr_number>

Contributors (optional):

@yangfanglin @JongilHan66

[lisa-bengtsson]

This looks like a lot of (important) work - just a heads up that I just committed a PR updating progomega_calc.f90 - looks like we should be able to merge the two in a straight forward way. #383
We can see which one goes first and then decide how to proceed.

[rhaesung]

This looks like a lot of (important) work - just a heads up that I just committed a PR updating progomega_calc.f90 - looks like we should be able to merge the two in a straight forward way. #383
We can see which one goes first and then decide how to proceed.

@lisa-bengtsson Thanks for the heads-up! Since this PR is currently in draft status while we conduct 32-bit scientific validation, we can proceed with #383 first. Once it is merged, I will rebase and apply the sp implementation to the updated code.

[lisa-bengtsson]

This looks like a lot of (important) work - just a heads up that I just committed a PR updating progomega_calc.f90 - looks like we should be able to merge the two in a straight forward way. #383
We can see which one goes first and then decide how to proceed.

@lisa-bengtsson Thanks for the heads-up! Since this PR is currently in draft status while we conduct 32-bit scientific validation, we can proceed with #383 first. Once it is merged, I will rebase and apply the sp implementation to the updated code.

@rhaesung , great! Thank you!