WRF selector 31 forcing source notice

WRF was developed at the National Center for Atmospheric Research (NCAR) which is operated by the University Corporation for Atmospheric Research (UCAR). NCAR and UCAR make no proprietary claims, either statutory or otherwise, to this version and release of WRF and consider WRF to be in the public domain for use by any person or entity for any purpose without any fee or charge. UCAR requests that any WRF user include this notice on any partial or full copies of WRF. WRF is provided on an "AS IS" basis and any warranties, either express or implied, including but not limited to implied warranties of non-infringement, originality, merchantability and fitness for a particular purpose, are disclaimed. In no event shall UCAR be liable for any damages, whatsoever, whether direct, indirect, consequential or special, that arise out of or in connection with the access, use or performance of WRF, including infringement actions.

WRF® is a registered trademark of the University Corporation for Atmospheric Research (UCAR).

Source: https://github.com/wrf-model/WRF/blob/v4.7.1/LICENSE.txt

The forcing sources have identical SHA-256 hashes in WRF 4.6.1 and 4.7.1:
phys/module_ra_hs.F
  7a83582205dedd31071a8fbb3307e227be7e244f282db278c84368e6ae232e3e
dyn_em/module_damping_em.F
  933628eae4ef09513d338317529d252933ef602c8659972b99ad67d70583c98f

No coefficient table files are required. Source headers carry no additional
licence restriction. The existing float32 math helpers retain their own
permissive notices in the kernel headers and packaged licences.

The initial zero surface-longwave convention was also checked against:
https://raw.githubusercontent.com/wrf-model/WRF/v4.6.1/phys/module_radiation_driver.F
  779875651512709c47d50ef6c1e1216ad94726f1996c5e45a0db8eb800733ffa
This file is part of the same public-domain WRF distribution. Its GLW
clear precedes the HSRAD dispatch; this scheme computes no surface flux.

The global ideal building blocks were checked at source against the same
distribution notice, before transcription. No source header adds a licence
restriction and none requires a coefficient table:
dyn_em/module_initialize_heldsuarez.F (identical in 4.6.1 and 4.7.1)
  f93af2b6e3303af3efb92dc119decd8cf3bd60489d5a12845a00041f45167c33
dyn_em/module_polarfft.F (identical in 4.6.1 and 4.7.1)
  12d62a8493f78e09a68aa6a216facebea90f4b456c800b62a605010813754d1e
dyn_em/module_big_step_utilities_em.F (4.6.1, calc_p_rho_phi)
  8f0649b458fceabd5c31c87ec9f266840fb164e2629322904a3f6ee04c72331a

The GPU polar transform calls the installed CUDA toolkit's cuFFT binary64
batched API through CuPy. No cuFFT binary or proprietary source is vendored.
The shipped weighting, packing and writeback kernels are derived from WRF.
The required CUDA toolkit is already a platform runtime dependency.

dyn_em/module_small_step_em.F (4.6.1, hydrostatic calc_p_rho)
  cabf1a177d50fb0096db79644af20cfe6d75217dbe63ab406a7e29bb54c17634
Same WRF distribution notice; source inspected before transcription.

The directional Coriolis and polar curvature operators also derive from
the pinned module_big_step_utilities_em.F above. Global advection derives
from the same public-domain distribution, inspected before transcription:
dyn_em/module_advect_em.F (4.6.1, advect_scalar/u/v)
  58253bdbeb188dd47ed0579fcd2891086be1889b75c0c7d3696c9ad1d213559d

The global schedule follows the same public-domain WRF distribution:
dyn_em/solve_em.F (4.6.1)
  e42df5d7db4b6ec4a3b8e2f228a8ec8f9a4c426656093bfcebe58a8de6c3e8f4
The unedited stock namelist is retained as tests/data/global_heldsuarez.input:
test/em_heldsuarez/namelist.input (4.6.1)
  349086813259f42c38714477d9bc94ae4b61b54fc2a9f7ca05c7f187fee786f2
Both sources were checked under LICENSE.txt above. Neither source adds a
licence restriction or requires an external coefficient table.
