Metadata on 5x5 WISE mapunits dataset is misleading

[ekluzek]

The metadata on the 5x5 WISE mapunits soil dataset is misleading and should be corrected...

It looks like a surface dataset rather than an input dataset to mksurfdata.

ncdump -h $CSMDATA/lnd/clm2/rawdata/mksrf_soil_mapunits_5x5min_WISE.c220330.nc
netcdf mksrf_soil_mapunits_5x5min_WISE.c220330 {
dimensions:
	lsmlat = 2160 ;
	lsmlon = 4320 ;
variables:
	double LATIXY(lsmlat, lsmlon) ;
		LATIXY:long_name = "latitude" ;
		LATIXY:units = "degrees north" ;
	double LONGXY(lsmlat, lsmlon) ;
		LONGXY:long_name = "longitude" ;
		LONGXY:units = "degrees east" ;
	int MU(lsmlat, lsmlon) ;
		MU:long_name = "soil texture map units" ;
		MU:units = "unitless" ;

// global attributes:
		:Conventions = "NCAR-CESM" ;
		:History_Log = "created on: 03-30-22 13:20:53" ;
		:Source = "Community Land Model: CLM5" ;
		:Version = "ctsm5.1.dev077-122-gc79e8fc72" ;
		:Logname = "mvertens" ;
		:Host = "cheyenne5" ;
		:Input_grid_dataset = "UNSTRUCTgrid_5x5min_nomask_c200129.nc" ;
		:Inland_lake_raw_data_file_name = "mksrf_LakePnDepth_3x3min_simyr2004_csplk_c151015.nc" ;
		:Inland_wetland_raw_data_file_name = "mksrf_lanwat.050425.nc" ;
		:Glacier_raw_data_file_name = "mksrf_glacier_3x3min_simyr2000.cdf5.c120926.nc" ;
		:Glacier_region_raw_data_file_name = "mksrf_GlacierRegion_10x10min_nomask_cd5_c220131.nc" ;
		:Urban_Topography_raw_data_file_name = "mksrf_topo.10min.cdf5.c220201.nc" ;
		:Urban_raw_data_file_name = "mksrf_urban_0.05x0.05_simyr2000.c220127.nc" ;
		:Vegetation_type_raw_data_filename = "mksrf_landuse_histclm50_LUH2_2000.cdf5.c170629.nc" ;
		:agfirepkmon_raw_data_file_name = "mksrf_abm_0.5x0.5_AVHRR_simyr2000.c130201.nc" ;
		:gdp_raw_data_file_name = "mksrf_gdp_0.5x0.5_AVHRR_simyr2000.c130228.nc" ;
		:peatland_raw_data_file_name = "mksrf_peatf_0.5x0.5_AVHRR_simyr2000.c130228.nc" ;
		:soildepth_raw_data_file_name = "mksf_soilthk_5x5min_ORNL-Soil_simyr1900-2015_c170630.nc" ;
		:topography_stats_raw_data_file_name = "mksrf_topostats_1km-merge-10min_HYDRO1K-merge-nomask_simyr2000.c130402.nc" ;
		:mesh_pft_file_name = "UNSTRUCTgrid_0.25x0.25_nomask_c200129.nc" ;
		:mesh_lakwat_file = "UNSTRUCTgrid_3x3min_nomask_c200129.nc" ;
		:mesh_wetlnd_file = "UNSTRUCTgrid_0.5x0.5_nomask_c200129.nc" ;
		:mesh_glacier_file = "UNSTRUCTgrid_3x3min_nomask_c200129.nc" ;
		:mesh_glacier_region_file = "UNSTRUCTgrid_10x10min_nomask_cdf5_c200129.nc" ;
		:mesh_soil_texture_file = "UNSTRUCTgrid_5x5min_nomask_c200129.nc" ;
		:mesh_soil_color_file = "UNSTRUCTgrid_0.25x0.25_nomask_c200129.nc" ;
		:mesh_soil_organic_file = "UNSTRUCTgrid_5x5min_nomask_c200129.nc" ;
		:mesh_urban_file = "UNSTRUCTgrid_3x3min_nomask_c200129.nc" ;
		:mesh_fmax_file = "UNSTRUCTgrid_0.125x0.125_nomask_c200129.nc" ;
		:mesh_VOC_EF_file = "UNSTRUCTgrid_0.5x0.5_nomask_c200129.nc" ;
		:mesh_harvest_file = "UNSTRUCTgrid_0.25x0.25_nomask_c200129.nc" ;
		:mesh_urban_topography_file = "UNSTRUCTgrid_10x10min_nomask_c200129.nc" ;
		:mesh_agfirepkmon_file = "UNSTRUCTgrid_0.5x0.5_nomask_c200129.nc" ;
		:mesh_gdp_file = "UNSTRUCTgrid_0.5x0.5_nomask_c200129.nc" ;
		:mesh_peatland_file = "UNSTRUCTgrid_0.5x0.5_nomask_c200129.nc" ;
		:mesh_soildepth_file = "UNSTRUCTgrid_5x5min_nomask_c200129.nc" ;
		:mesh_topography_stats_file = "UGRID_1km-merge-10min_HYDRO1K-merge-nomask_c130402.nc" ;
		:lai_raw_data_file_name = "mksrf_lai_78pfts_simyr2005.cdf5.c220206.nc" ;
		:soil_color_raw_data_file_name = "mksrf_soilcolor_CMIP6_simyr2005_cd5.c220126.nc" ;
		:soil_texture_raw_data_file_name = "mksrf_soil_mapunits_30sec_WISE.c220330.nc" ;
		:fmax_raw_data_file_name = "mksrf_fmax_0.125x0.125_c200220.nc" ;
		:organic_matter_raw_data_file_name = "mksrf_organic_10level_5x5min_ISRIC-WISE-NCSCD_nlev7_c120830.nc" ;
		:VOC_EF_raw_data_file_name = "mksrf_vocef_0.5x0.5_simyr2000.c110531.nc" ;
		:history = "Wed Mar 30 16:06:53 2022: ncrename -v mapunits,MU soiltex_mapunits_4320x2160_c220329.nc\n",
			"Wed Mar 30 14:44:57 2022: ncks -v mapunits,LATIXY,LONGXY surfdata_4320x2160_hist_78pfts_CMIP6_2000_c220329.nc soiltex_mapunits_4320x2160_c220329.nc" ;
		:NCO = "netCDF Operators version 4.9.5 (Homepage = http://nco.sf.net, Code = http://github.com/nco/nco)" ;
}

[wwieder]

Thanks for this note, @ekluzek .

I'm fuzzy on the details but the file I created and handed off to @mvertens was created by this script. I recall @mvertens then had to add some metadata to get mksrfdata_esmf to work, but maybe not all of this information was needed?

I'm assuming it's just some of the global attributes that look odd? Are there other data (maybe from my intermediate file, that's not been scrubbed) that should have been retained?

At this point I don't want to spend a ton of time on this, but it would be good to clean up our house related to documentation before the CESM3 release.

[wwieder]

If you're interested, the 'raw data' I generated from the orig. ISRIC-WISE tiff file is here

$ncdump -h /glade/campaign/cgd/tss/people/wwieder/obs_dataset/ISRIC-WISE/WISE30sec/Interchangeable_format/wise_30sec_v1_grid2.nc
netcdf wise_30sec_v1_grid2 {
dimensions:
	lon = 43200 ;
	lat = 16753 ;
	bound = 2 ;
variables:
	double lon(lon) ;
		lon:_FillValue = 1.e-36 ;
		lon:long_name = "longitude" ;
		lon:standard_name = "longitude" ;
		lon:units = "degrees_east" ;
		lon:bounds = "lon_bnds" ;
	double lat(lat) ;
		lat:_FillValue = 1.e-36 ;
		lat:long_name = "latitude" ;
		lat:standard_name = "latitude" ;
		lat:units = "degrees_north" ;
		lat:bounds = "lat_bnds" ;
	int MU(lat, lon) ;
		MU:_FillValue = 0 ;
		MU:scale_factor = 1. ;
		MU:add_offset = 0. ;
		MU:long_name = "WISE30sec MapUnit" ;
	int bound(bound) ;
	double lon_bnds(lon, bound) ;
		lon_bnds:_FillValue = 1.e-36 ;
	double lat_bnds(lat, bound) ;
		lat_bnds:_FillValue = 1.e-36 ;

// global attributes:
		:sourceData = "https://files.isric.org/public/wise/wise_30sec_v1.zip" ;
		:doucmentation = "https://library.wur.nl/WebQuery/wurpubs/fulltext/400244" ;
		:publication = "https://doi.org/10.1016/j.geoderma.2016.01.034" ;
		:accessed = "11 March, 2022" ;
		:notes = "generated by Will Wieder" ;
		:sourceCode = "ctsm_py/notebooks/tiff2nc-WISE.ipynb" ;

[ekluzek]

@wwieder that's what I think as well, this is low priority, but would be good to fix before CESM3. And it's just the global attributes that need to be updated. Largely the attributes from the 30-sec file above should replace what's currently on the 5-minute one. But, there also needs to be something about how the data was averaged up to 5-min. I assume it was something like the dominant one in the 5-min grid cell?

Other than the global attributes I don't think any of the other data on the 30-sec file needs to be moved to the 5-min one. And the meta-data on the 5-min fields seems fine to me.