Add spatially heterogeneous soil colour maps for NoahMP (based on feature/monan-757-NF)#19
Add spatially heterogeneous soil colour maps for NoahMP (based on feature/monan-757-NF)#19mpaiao wants to merge 23 commits into
Conversation
This is the first commit to implement geographically variable soil colour initialisation in MONAN. First, the original Noah-MP table of albedo thresholds was replaced by the CLM-5, and the code was edited so the soil colour class is no longer hardcoded.
…and replaces the original classes in NoahMP (8 classes) with those defined in CLM-5 (20 classes). This has not been tested yet, so it is likely the first of a few commits until the new capability is fully functional.
mpaiao
changed the title
There was a problem hiding this comment.
Marcos, conforme definimos em conjunto na reunião de hoje (03/06/26), segue as seguintes necessidades:
- Alteração no Pull Request visando isolar e manter a autonomia de execução tanto do noahmp original quanto do noahmp alterado com soilcolour_30s (para o MONAN), podendo via namelist.atmosphere.TEMPLATE ativar ou não essa opção.
- Caso o usuario opte por ativar essa opção no namelist (acrescentando a linha config_lsm_scheme = 'sf_noahmp_monan' por exemplo) nesse caso o modelo utilizaria os arquivos de soilcolour_30s com a tabela NoahmpTable.TBL alterada para 21 categorias de cor de solo.
- Caso o usuário não altere nada, o MONAN segue sua execução normal utilizando os arquivos de entrada default.
Obrigado
GCC
…o finish reverting it.
…ants the default settings and has not updated NoahmpTable.TBL
… colour directory is not found. This __should__ allow for the code to proceed without failing.
…outine. It seems the code is overwriting the default value.
Now it must always be "MODIFIED_RAD_CLM_NOAH".
|
[English version below] Oi @netosylvio, eu atualizei o Pull Request. Fiz um pouco diferente do que havíamos discutido, seguindo uma sugestão do @marcelo-santini, que achei mais elegante. Ao invés de duplicar o A diferença de lógica entre a vegetação e cor do solo é que, no caso da cor do solo, se o(a) usuário(a) não configurar o MONAN explicitamente para usar cor de solo heterogênea, o código vai usar apenas os namelists que já existiam, e o código vai rodar também com cor de solo homogênea que já existia no NoahMP para o domínio todo. Eu testei as duas configurações na Jaci, e ambas estão funcionando como esperado. E atualizei o comentário inicial para explicar melhor estas atualizações. Obrigado pelo feedback, e desculpe-me pela demora em ver isto! Hi @netosylvio, I updated this Pull Request. I refactored it slightly differently from our previous discussions, following a suggestion from @marcelo-santini that I think is neater. Instead of duplicating The difference between the vegetation and soil colour logics is that, for soil colour, the code will fall back to defaults in case the user does not explicitly requests the new heterogeneous soil colour maps. In this case, the code will just search for the namelists that already existed in I tested both settings on Jaci, and both seem to be working as expected. I updated my opening post to explain how to enable the spatially heterogeneous soil albedo parameters. Thanks for the feedback, and apologies for taking long to get back to this. |
3 participants
Pull Request Description
This pull request supersedes PR #18 as it is based on branch
feature/monan-757-NF.Updates in this pull request include (1) replacement of the original soil colour categories from NoahMP with those from CLM-5 and (2) soil colour initialisation in NoahMP based on input maps. These changes addresses the issue that native NoahMP simulations have too little variation in reflectivity across sparsely vegetated regions known to have marked variation in albedo (e.g., Sahara, Namibian desert).
Most changes follow the steps to initialise soil texture in NoahMP, and were needed to ensure each grid cell sets the reflectivity parameter based on the input data. This pull request does not provide any data itself: if these data sets are not provided, the code will fall back to default (i.e., globally homogeneous soil colour index). By default, this option is disabled.
For users not interested in running with heterogeneous soil colour, no action is needed. Otherwise, they must follow these steps:
geog_sub_path), under a path namedclm_soilcolour_21class_30s. Scripts that read in CLM-5 data and translate them to NoahMP format are available at this git repository. Or if running MONAN on Jaci, make sure to point to a path where this is already available.namelist.init_atmosphere, edit namelistdata_sourcesto include the following:config_soilcol_data = 'DEFAULT_RAD_NOAH')config_soilcol_datais used during the model execution. If usingscripts_CD-CT, make sure that bothnamelist.init_atmosphere.TEMPLATEandnamelist.init_atmosphere.STATIChave this configuration, and make sure that the script overwrites the static files under step 2.NoahmpTable.TBLas provided in the source code (/path/to/MONAN-Model/src/core_atmosphere/physics/physics_noahmp/parameters/NoahmpTable.TBL).Collaborators
@aomanzi @marcelopaivaramos @carlosrenatosouza2 @pkubota @marcelo-santini
Type of Change
Testing and Quality
Scientific Impact
This implementation should allow for better spatial representation of surface reflectivity at global scales. This is simply a rebase of PR #18 onto
feature/monan-757-NF. The results are consistent with the previous pull request. The differences in albedo (defined in the bulk form, i.e., ratio of average upwelling solar irradiance and downwelling solar irradiance, both integrated over a 48-hour period) between this pull request and the baseline version (feature/monan-757-NF).