milancurcic
New member
Hi All,
I have a few questions about correctly preparing geo_em*.nc and wrfinput* files for running WRF with the BEP urban scheme (sf_urban_physics = 2), both with single-urban-class and 3-urban-classes scenarios.
The urban fraction and classes are derived from the 2015 30-m Landsat data for North America (http://www.cec.org/north-american-environmental-atlas/land-cover-30m-2015-landsat-and-rapideye/). Within WPS and WRF, I'm working with the MODIFIED_IGBP_MODIS_NOAH table.
My procedure is:
a) Run geogrid.exe to generate geo_em.d0[123].nc. I provide NLCD2011 and NUDAPT44 data at this stage. The resulting files inclued non-zero FRC_URB2D and URB_PARAM fields.
b) Extract urban fractions from the 2015 30-m TIFF data. I update FRC_URB2D in geo_em.d0[123].nc. I also update LU_INDEX to values 31, 32, and 33 for grid cells where I identify 3 urban classes (low density residential, high density residential, and commercial/industrial).
c) Run metgrid.exe to generate met_em.*.nc.
d) Run real.exe to generate wrfinput_d0[123] and wrfbdy_d01.
My questions are:
1. When I run real.exe with num_land_cat=21 (not adjusting for classes 31, 32, 33 that I added), I see that the resulting wrfinput_d0[123] have all urban multi-classes (what used to be 31, 32, 33 in geo_em* and met_em*) are now collapsed into a single urban class 13. However, if I set num_land_cat=33 and run real.exe, urban multi-classes are again collapsed into 13, and all other (non-urban) classes in the domain are equal to 1. That can't be correct, right? What is the correct way to run real.exe here to preserve urban multi-classes from met_em* files? Do I perhaps need to modify LU_INDEX in wrfinput* files rather than geo_em* files?
2. How does BEP determine where are the 3 urban classes. Is it from LU_INDEX or from FRC_URB2D? If the latter, should I not bother with multi-classes in LU_INDEX?
3. How to correctly run BEP with a single urban class? Do I need to modify URBPARM.TBL to include parameters only for one class?
Thank you!
Milan
I have a few questions about correctly preparing geo_em*.nc and wrfinput* files for running WRF with the BEP urban scheme (sf_urban_physics = 2), both with single-urban-class and 3-urban-classes scenarios.
The urban fraction and classes are derived from the 2015 30-m Landsat data for North America (http://www.cec.org/north-american-environmental-atlas/land-cover-30m-2015-landsat-and-rapideye/). Within WPS and WRF, I'm working with the MODIFIED_IGBP_MODIS_NOAH table.
My procedure is:
a) Run geogrid.exe to generate geo_em.d0[123].nc. I provide NLCD2011 and NUDAPT44 data at this stage. The resulting files inclued non-zero FRC_URB2D and URB_PARAM fields.
b) Extract urban fractions from the 2015 30-m TIFF data. I update FRC_URB2D in geo_em.d0[123].nc. I also update LU_INDEX to values 31, 32, and 33 for grid cells where I identify 3 urban classes (low density residential, high density residential, and commercial/industrial).
c) Run metgrid.exe to generate met_em.*.nc.
d) Run real.exe to generate wrfinput_d0[123] and wrfbdy_d01.
My questions are:
1. When I run real.exe with num_land_cat=21 (not adjusting for classes 31, 32, 33 that I added), I see that the resulting wrfinput_d0[123] have all urban multi-classes (what used to be 31, 32, 33 in geo_em* and met_em*) are now collapsed into a single urban class 13. However, if I set num_land_cat=33 and run real.exe, urban multi-classes are again collapsed into 13, and all other (non-urban) classes in the domain are equal to 1. That can't be correct, right? What is the correct way to run real.exe here to preserve urban multi-classes from met_em* files? Do I perhaps need to modify LU_INDEX in wrfinput* files rather than geo_em* files?
2. How does BEP determine where are the 3 urban classes. Is it from LU_INDEX or from FRC_URB2D? If the latter, should I not bother with multi-classes in LU_INDEX?
3. How to correctly run BEP with a single urban class? Do I need to modify URBPARM.TBL to include parameters only for one class?
Thank you!
Milan