New urban variables from BEP-BEM for the FPS-URB-RCC and improve memory usage

[jesusff]

This PR incorporates Alberto Martilli's modifications to save CORDEX FPS-URB-RCC variables, improve memory usage, and fix bugs in Noah and Noah-MP parts of the code related to urban parametrizations. The base version is v4.5.1.6.

TYPE

new features, bug fixes

KEYWORDS

urban, BEP+BEM, Noah-MP, Noah

SOURCE

Alberto Martilli

DESCRIPTION OF CHANGES

Problems:

• WRF crashes when using the BEP+BEM urban parameterization due to excessively large wrfrst files. A temporary workaround is to set ioform_rst = 102 in the namelist.input, which is not a practical solution for long-term simulations.
• The BEP+BEM urban parametrization scheme is unable to output variables within the urban canopy.

Solution:

• Temporary solution: parameter nbui_max decreased and dz_u increased from 5 to 10m in module_sf_bep_bem.F
• New variables added: TSK_RURAL, T2R_URB2D, T2G_URB2D, TGROUND_URB2D, TROOF_URB2D, T2VEG_URB2D, TCAN_URB2D, UCAN_URB2D, QCAN_URB2D

Bugs fixed:

• When using LCZ classifications, NoahMP does not accurately compute near-surface diagnostic variables for urban grid cells.
• There is an incorrect calculation of T2VEG_URB2D in the Noah LSM.

New additions:

• Formulation following Kanda et al. (2007) added in module_sf_bep_bem.F to estimate roughness length for heat, as used in the SLUCM.

LIST OF MODIFIED FILES:

Registry.EM_COMMON
module_first_rk_step_part1.F
module_initialize_real.F
module_sf_bep_bem.F
module_surface_driver.F
module_sf_noahmpdrv.F
module_sf_noahdrv.F

TESTS CONDUCTED:

The updated version has been tested with all modifications applied, including a 2-day run at 3 km resolution over Buenos Aires and a 7-day run over Paris. Sensitivity tests to various parameters were conducted. No strong sensitivity was found for nbui_max and dz_m. The size of the wrfrst files has been significantly reduced, which make long-term runs with BEP+BEM scheme feasible.

Urban model near-surface diagnostics are now extracted, and standard near-surface diagnostics have been corrected. For the latter, additional tests over Madrid were performed, showing a difference of approximately 0.5 K in near-surface temperature (T2) compared to the previous version without the Noah-MP bug fix. The new version also shows improved agreement with observations.

RELEASE NOTE:

BEP+BEM features now include the extraction of urban canopy variables and the calculation of standard near-surface variables over LCZ grids is fixed for Noah-MP.

[jesusff]

@yoselita , I will merge so we can keep working on the latest version but, when you have time, comment on the modifications included, so we can use it in the RELEASE NOTE

Thanks!

[yoselita]

v4.5.1.7-devel until now should:

1. started from WRFv4.5.1-cwc
2. contain all modifications introduced by AMartilli
3. contain all bug fixes done on bug-fixes branch (Commits done on May 9, 2025)
4. contain additional 2 modification introduced by AMartilli (commits a03eaf8 and 3d3c973)

To my understanding, points 3 and 4 are not included yet. To do that, branch bug-fixes should be merged to v4.5.1.7-devel. This refers to pull request #4. Is this correct @jesusff?

[jesusff]

Hi,

yes, but I meant to have a brief text explanation of the modifications in your point 2 above, which is what this PR (branch urb-vars) contained. Thanks!