Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

update the urban land use land cover using a TIF file

arman

New member
Hi all,

I need to run WPS for Canada and update the urban land use land cover using a TIF file from the 2020 Land Cover of Canada. I can use GDAL to translate the TIF file to binary files, but I must remap the numbers first. The MODIFIED_IGBP_MODIS_NOAH classification scheme, I can find in the list there is number 13 for urban and built-up and 31, 32, and 33 for LOW_DENSITY_RESIDENTIAL, HIGH_DENSITY_RESIDENTIAL, and HIGH_INTENSITY_INDUSTRIAL. I don't have information for 31, 32, and 33 for LOW_DENSITY_RESIDENTIAL, HIGH_DENSITY_RESIDENTIAL, and HIGH_INTENSITY_INDUSTRIA in the new land-use file. Can I remap them as urban pixels to 13?

Thanks,
 
Thanks, Ming. I am a little bit confused by the following steps. I generated the bil file, and this is the header generated using gdal (please see below). What are the next steps I have to take? I can not find a straightforward guideline with which to proceed. Please give me a link that discusses the steps, or write me about what I must do afterward. Thanks

ENVI
description = {
data.bil}
samples = 190001
lines = 160001
bands = 1
header offset = 0
file type = ENVI Standard
data type = 1
interleave = bsq
byte order = 0
map info = {Lambert Conformal Conic, 1, 1, -2600010, 3914940, 30, 30}
projection info = {4, 6378137, 6356752.314140356, 49, -95, 0, 0, 49, 77, Lambert Conformal Conic}
coordinate system string = {PROJCS["NAD_1983_CSRS_Canada_Atlas_Lambert",GEOGCS["GCS_North_American_1983_CSRS",DATUM["D_North_American_1983_CSRS",SPHEROID["GRS_1980",6378137.0,298.257222101]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]],PROJECTION["Lambert_Conformal_Conic"],PARAMETER["False_Easting",0.0],PARAMETER["False_Northing",0.0],PARAMETER["Central_Meridian",-95.0],PARAMETER["Standard_Parallel_1",49.0],PARAMETER["Standard_Parallel_2",77.0],PARAMETER["Latitude_Of_Origin",49.0],UNIT["Meter",1.0]]}
band names = {
Canada2020}
 
Would you please take a look at this post, which discuss a similar issue that you raised. Hope you can find helpful information there.
 
Thanks Ming, it was helpful. In the post you shared with me, he didn't noted if he used the step 3 in manual which is proposed to to extract the urban categories from the binary tile and write a new tile of data containing urban fraction (copied below). As I noted in above I used 2020 Land Cover of Canada. Do I need to do step 3 as well? When i click on converter (this converter) in the manual, I have two options 1- For converting (x,y) coordinates in an Albers equal-area projection to a (lat,lon) location, pleaseuse the convert_albers code. 2- To extract urban categories and write an urban fraction tile, please use the uf code. Do I need to use both options?

I confused about using this step since the canadian seems doesnt have all required information for this step. For example if I want to use "uf code", I need to adjust the urban fractions for low-density, high-density, and commercial/industrial areas. For example some thing like this: ./uf -L 40 -H 80 -I 90 data2.bil.
Would you please advise if step 3 should be used. Thanks

3- Use this converter program to extract the urban categories from the binary tile and write a new tile of data containing urban fraction. The output file of this converter should be copied over the original land use tile, i.e., the urban fraction file should be renamed to 00001-ncols.00001-nrows, where ncols is the number of columns (in i5.5 format) and nrows is the number of rows (also in i5.5 format) in the tile, as in Step 2.
 
Arman,

Personally I think you can skip Step 3. In most cases we only offer urban landuse type and don't worry for the low-density, high-density, and commercial/industrial areas. But I am not 100% sure whether I am right. Please try and see whether it is fine to skip Step 3 in your case.
 
Hi Ming,

Thanks for your reply. I got the error below from geogrid.exe. I used 40 nodes and 80 CPUs to run geogrid.exe, and I think I have enough memory. I attached the mpi_output containing the error and allocated memory, geogrid.log, and the index file I created for the new binary file. I used North American Land Cover, 2024 (Landsat, 30m) since it was the file recommended in the WRF manual. I also copied the part I added to geogrid.tbl file below. I followed all the steps in the manual. Can you please let me know how I can solve this error? Thanks.


forrtl: severe (41): insufficient virtual memory
Image PC Routine Line Source
geogrid.exe 0000000000628043 Unknown Unknown Unknown
geogrid.exe 00000000004AAE86 Unknown Unknown Unknown
geogrid.exe 000000000047C3F2 Unknown Unknown Unknown
geogrid.exe 0000000000471624 Unknown Unknown Unknown
geogrid.exe 0000000000470450 Unknown Unknown Unknown
geogrid.exe 00000000004642A3 Unknown Unknown Unknown
geogrid.exe 0000000000408B6C Unknown Unknown Unknown
geogrid.exe 00000000004071D2 Unknown Unknown Unknown
libc-2.17.so 00002B133F01A555 __libc_start_main Unknown Unknown
geogrid.exe 00000000004070E9 Unknown Unknown Unknown


===============================
name = FRC_URB2D
priority=2
dest_type=continuous
fill_missing = 0.
interp_option=default: average_gcell(1.0)+four_pt
abs_path=default:/gpfs/fs0/scratch/m/mhatzo/ganjiarm/Models/MainModel/WRFmodel/wrf_CanUrbanLand/land_cover_2020v2_30m_tif/urban_fraction/
===============================
 

Attachments

  • mpi_output_14003125.txt
    6.3 KB · Views: 0
  • geogrid.log.0000
    25.2 KB · Views: 1
  • index.txt
    400 bytes · Views: 1
Hi Ming,

After looking at the urban fraction data, I encountered an issue with geo_em files generated with geogrids. Below are the details, along with supporting plots for reference.
  • Left Plot: This is the urban fraction visualized in ArcGIS Pro, correctly located and aligned as expected.
  • Issue in WPS: While WPS successfully reads the urban fraction file, it seems to include the urban fraction multiple times, overlaying them on top of each other. This results in an incorrect representation of the data in the output geogrid. At least I can say that the generated tile was misplaced with geogrids.exe. The projection of bil file is Canada Albers Equal Area Conic_nad83. The selected projection in the index file is albers_nad83. I have to note that I followed all steps noted in manual : The WRF Preprocessing System (WPS) — WRF Users Guide documentation
To help diagnose the problem, I have attached the following files:
  1. Index File.
  2. GEOGRIDS.tbl
  3. Namelist File
  4. bil file
  5. geo_em file
Do you have any insights or suggestions on how to address this issue? Your expertise would be greatly appreciated.
.
1736955623932.png
 

Attachments

  • namelist.wps
    1.2 KB · Views: 1
  • question.zip
    7.4 MB · Views: 2
Last edited:
Top