Using CORINE land uses

Dataset specific topics involving WPS.

Using CORINE land uses

Postby darnold » Thu Jul 01, 2010 4:08 am

Dear Wrf users,

we want to use CORINE land uses in our wrf simulations for an European region. For this, not only the CORINE data itself must be transformed to an appropriate format for wps to read in, but also a new or adapted LANDUSE.TBL should be created with the proper albedo, emissivity, thermal inretia roughness and soil moisture.

When looking around in the literature you can find the very interesting paper by Pineda et. al 2004 which, for a specific are in the north east of Spain, used the AVHRR to get all the parameters but soil moisture and roughness, which were extrapolated from the NDVI vegetaton index. However, i would like to know if any other users have already implemented corine in WRF and if so, how did they generate the appropriate landuse table and, if possible, maybe provide su with the LANDUSE.TBL used for such matter.

Thanks.
darnold
 
Posts: 13
Joined: Wed Feb 17, 2010 5:28 am

Re: Using CORINE land uses

Postby darnold » Mon Nov 15, 2010 10:07 am

Dear Wrf users,

at the moment, follfowing Pineda et al. 2004 we have remapped CORINE land uses to the 24 LANDUSES available in the old MM5 (WRF has some additional LANDUSES) but without changing any of the default parameters liek moisture and roughness. This should allow, at least, to have a better representation of the changes in the soil from the 80ies until now (like, having larger cities and so on) since the spatial resolution is better.

This has been achieved using ARCGIS and creating a single raster file with a header like :

ncols 35105
nrows 29181
xllcorner -3.8659300236093
yllcorner 35.80246253386
cellsize 0.000833333
NODATA_value -9999


followed buy the rows of data.

In principle this should be easily arranged using write_geogrid.c routien and a piece of code as the one user lgiovannini kindly posted for us in post viewtopic.php?f=22&t=481&start=0 which nicely worked for us to use the srtm3" elevation data. We have modifed the .TBL and index files accordingly, however, when trying it with the landuses, the file written out is incorrect and when i "od" i see stramge values and running WPS gives errors such as "landuse 216 not found" , which obviously shouldn't be there.

Maybe one of theusers could give us a hint on how to proceed, even if the data used is not CORINE.

Many thanks!
Delia
darnold
 
Posts: 13
Joined: Wed Feb 17, 2010 5:28 am

Re: Using CORINE land uses

Postby darnold » Mon Apr 18, 2011 7:51 am

We have been asked some times how we use corine data in our runs. Therefore we have thought it could be useful to post a kind of very simple HOWTO here. So here it goes:

How to use CORINE in WPS/WRF
************************************************
for referencing:
Arnold, D., Schicker, I. and P. Seibert, 2010:
High-Resolution Atmospheric Modelling in Complex Terrain for Future Climate Simulations(HiRmod). VSC Report 2010
http://www.boku.ac.at/met/envmet/hirmod.html
delia.arnold[at]boku.ac.at
irene.schicker[at]boku.ac.at
************************************************

The procedure is similar to using srtm data (viewtopic.php?f=22&t=481). It has two basic steps:

1. Remap, reproject and convert the original CORINE 2000/2006 tif file to ASCII
2. Convert to WPS-binary readable

=== REMAP/REPROJECT/CONVERT (ARCGIS needed) ===
- Download CLC data from http://www.eea.europa.eu
- Extract all the data in one directory
- Open ArcGIS and load the tif file extracted
- Remap: Take Table 1 in Pineda et al. 2004 and use it to reclassify the new classes into the USGS ones with the ArcGIS Spatial Analyst Toolbox
(under Reclass) -- maybe better on a step-wise approach due to memory requirements.
- Reproject: reprojection into WGS84 with the ArcGIS Data Management Tools and tool Projections transformations -> Raster-> Project Raster.
- Convert: to an ASCII file using the ArcGIS Conversion Tools -> From Raster -> Raster to Ascii

We get now an ASCII file with a header containing:
ncold XXXX
nrows XXXX
xllcorner XXXX.XXXXX
yllcorner XXXX.XXXXX
cellsize X.XXXXXXX
NODATA_value -9999

This information will be needed in the second step

References:
Pineda et al. 2004: N. Pineda, O. Jorba, J. Jorge, J.M. Baldasano. Using NOAA AVHRR and SPOT VGT data to estimate surface parameters: application to
a mesoscale meteorological model


=== Convert to WPS-binary readable format ===
(with a slightly modified version of the fortran code provided in viewtopic.php?f=22&t=481&start=10#p12411 by
Lorenzo Giovannini, Atmospheric Physics Group Department of Civil and Environmental Engineering, Universiy of Trento)

- (paths can be modified, just be sure to compile with the appropriate routines) in
WPS/geogrid/src create a program to read in the data and with the help of write_geogrid.o
transform it to WPS format. A sample program is:

*****wpsingest_toCLC.F*****

Code: Select all
program geogrid_clc
implicit none

integer :: i,j
integer :: isigned, endian, wordsize
integer :: nx, ny, nz
real :: scalefactor
real*8 :: xllcorner, yllcorner, cellsize, missvalue
character :: head12
real, allocatable :: rarray(:,:), iarray(:,:)

isigned = 1
endian = 0
wordsize = 2
scalefactor = 1.0
nz = 1


! read in the ascii new landuse data
open (10, file = 'file.asc')

!read in the header
read(10,*) head12, nx
read(10,*) head12, ny
read(10,*) head12, xllcorner
read(10,*) head12, yllcorner
read(10,*) head12, cellsize
read(10,*) head12, missvalue

allocate(rarray(nx,ny))
allocate(iarray(nx,ny))

!read in the data
do j = 1,ny
read(10,*) iarray(:,j)
end do

! reverse the data so that it begins at the lower-left corner
do j = 1,ny
rarray(:,j) = iarray(:,ny-(j-1))
enddo

!set the missing values
do j = 1, ny
do i = 1, nx
if ( rarray(i,j) < 0 ) then
rarray(i,j) = -99 ! set negative terrain to be zero since those are near coastal or river
banks
end if
end do
end do


call write_geogrid(rarray, nx, ny, nz, isigned, endian, scalefactor, wordsize)

end program


- Compile the write_geogrid.c:
Code: Select all
pgcc -D_UNDERSCORE -DBYTESWAP -DLINUX -DIO_NETCDF-DIO_BINARY -DIO_GRIB1  -D_GEOGRID -O -c  write_geogrid.c


- Compile the ingestion program with the write_geogrid.o
Code: Select all
pgf90 -Mfree wpsingest_toCLC.F write_geogrid.o


- This will create the data in the tile format described in the WPS documentation.

- Copy/move/link this file to where the geog data is stored: /Data/geog/landuse_3s

- Create an index file using the header of the ASCII file created in step 1 similar to:
type=categorical
category_min=1
category_max=24
projection=regular_ll
dx=0.000833333
dy=0.000833333
known_x=1.0
known_y=1.0
known_lat=35.80246253386
known_lon=-3.8659300236093
wordsize=1
tile_x=35105
tile_y=29181
tile_z=1
missing_value = 16
units="category"
description="24-category USGS landuse-from corine"

**!!**If the missign values are defined as water, as above (16) mind the terrain will be
then be masked by
them!

- Edit the GEOGRID.TBL and add the entry for the new landuse!
===============================
name=LANDUSEF
priority=2
dest_type=categorical
z_dim_name=land_cat
landmask_water = default:16 # Calculate a landmask from this field
interp_option = 3s:nearest_neighbor
rel_path= 3s:landuse_3s/
===============================
name=LANDUSEF
priority=1
dest_type=categorical
z_dim_name=land_cat
landmask_water = modis_30s:17 # Calculate a landmask from this field
landmask_water = default:16 # Calculate a landmask from this field
dominant=LU_INDEX
interp_option = modis_30s:nearest_neighbor
interp_option = 30s:nearest_neighbor
interp_option = 2m:four_pt
interp_option = 5m:four_pt
interp_option = 10m:four_pt
interp_option = default:four_pt
rel_path= modis_30s:modis_landuse_20class_30s/
rel_path= 30s:landuse_30s/
rel_path= 2m:landuse_2m/
rel_path= 5m:landuse_5m/
rel_path= 10m:landuse_10m/
rel_path= default:landuse_2m/
===============================

and set in namelis.input:

geog_data_res = '3s+2m',
darnold
 
Posts: 13
Joined: Wed Feb 17, 2010 5:28 am

Re: Using CORINE land uses

Postby cycle_13 » Fri May 27, 2011 4:01 am

Dear everyone,
I am from east asia and also want using high resolution land use data into wrf model.From the website,http://www.eea.europa.eu, only can I download the Europe data ,which is not the area I am interested.So what I want to know is where I can find the land use data like CORINE data.
Thank you very much.
cycle_13
 
Posts: 27
Joined: Mon Jun 21, 2010 8:48 pm

Re: Using CORINE land uses

Postby julien1188 » Mon Sep 05, 2011 9:54 am

Thank you for your tutorial. It is interesting!

I followed each point but, unfortunatelly, I have a problem with CORINE landcover.

Firstly, I downloaded CORINE and ploted in arcGIS. I selected a zone with command "CLIP" of arcGIS. After, I transformed ETRS to WGS84.I converted raster to ascii. After, I convert ascii to bin with your sample program. But when I ploted LANDCOVER in geo_em.d02.nc, I had this plot (black line is coast line of my data visualization - ferret ): http://dl.dropbox.com/u/37896253/error.gif .If i plot this domaine in arcGIS, I have the correct zone (golf de gascogne). What's the problem? Projection?

This is the header of the ascii file:
Code: Select all
ncols         12525
nrows         9742
xllcorner     -8,0000639562732
yllcorner     41,999892475265
cellsize      0,00071857612541468
NODATA_value  -9999


and my index:
Code: Select all
type=categorical
category_min=1
category_max=24
projection=regular_ll
dx=0.00071857612541468
dy=0.00071857612541468
known_x=1.0
known_y=1.0
known_lat=41.999892475265
known_lon=-8.0000639562732
wordsize=1
tile_x=12525
tile_y=9742
tile_z=1
tile_bdr=0
missing_value = 0
units="category"
description="24-category USGS landuse-from 2006 CORINE (UE project)"


Thank you for your help
julien1188
 
Posts: 21
Joined: Sat Nov 06, 2010 9:43 am

Re: Using CORINE land uses

Postby darnold » Thu Sep 08, 2011 12:34 am

Hi,

sorry for the late reply, but i do not keep track of the posts (maybe a direct message is an easier way to get to me).

Please correct me if i am understanding right. You have plot the output of geogrid geo_em.d02.nc in ARCGIS and looks fine and in ferret no?

I am currently in a very active and passionate discussion on this and we are finding some issues regarding geoids used. However your differences in ferret are of 2 degrees!!! Is there any reprojection when mapping the data to be seen by ferret?

You could try also to plot your output with WTOOLS http://www.nusculus.com/wtools, it is a nice tool to plot your output in google earth.

Please tell me how it goes, shifts in topography are important issues for all us, the users!!!
darnold
 
Posts: 13
Joined: Wed Feb 17, 2010 5:28 am

Re: Using CORINE land uses

Postby PedroCorreia » Wed Feb 15, 2012 11:56 am

Hi,

first, thanks for the step by step manual. It was very helpful. I've managed to use SRTM (90m) and the Corine Land successfully in the last WRF version (3.3.1).

I just wanted to add some extra stuff i found to be needed (Under Linux):

- Convert Geotiff to asc (Both Corine and SRTM):
Code: Select all
gdal_translate -of “AAIGrid” $SRTM_FILE $File.asc


- Reclassify/Reproject Corine Land use to USGS
Install the open source GIS (grass) and do a r.reclass, using the Pineda et al. 2004 article as base.
Take in account that you need to obtain 24 Land use classes (1 to 24), and not (101 to 124).
Reproject the new raster with Lat/Lon projection into WGS84.

- Change the index and GEOGRID.TBL files:
GEOGRID.TBL
Code: Select all
===============================
name = HGT_M
        priority = 1
        dest_type = continuous
        df_dx=SLPX
        df_dy=SLPY
        smooth_option = smth-desmth_special; smooth_passes=1
        fill_missing=0.
   interp_option =      3s:four_pt
        interp_option =     30s:average_gcell(4.0)+four_pt+average_4pt
        interp_option =      2m:four_pt
        interp_option =      5m:four_pt
        interp_option =     10m:four_pt
        interp_option = default:four_pt
   rel_path=      3s:topo_3s/
        rel_path=     30s:topo_30s/
        rel_path=      2m:topo_2m/
        rel_path=      5m:topo_5m/
        rel_path=     10m:topo_10m/
        rel_path= default:topo_2m/
===============================
name = HGT_U
        output_stagger = U
        priority = 1
        dest_type = continuous   
   interp_option =      3s:four_pt
        interp_option =     30s:average_gcell(4.0)+four_pt+average_4pt
        interp_option =      2m:four_pt
        interp_option =      5m:four_pt
        interp_option =     10m:four_pt
        interp_option = default:four_pt
        smooth_option = smth-desmth_special; smooth_passes=1
        fill_missing=0.
   rel_path=      3s:topo_3s/
        rel_path=     30s:topo_30s/
        rel_path=      2m:topo_2m/
        rel_path=      5m:topo_5m/
        rel_path=     10m:topo_10m/
        rel_path= default:topo_2m/
===============================
name = HGT_V
        output_stagger = V
        priority = 1
        dest_type = continuous
   interp_option =      3s:four_pt
        interp_option =     30s:average_gcell(4.0)+four_pt+average_4pt
        interp_option =      2m:four_pt
        interp_option =      5m:four_pt
        interp_option =     10m:four_pt
        interp_option = default:four_pt
        smooth_option = smth-desmth_special; smooth_passes=1
        fill_missing=0.
   rel_path=      3s:topo_3s/
        rel_path=     30s:topo_30s/
        rel_path=      2m:topo_2m/
        rel_path=      5m:topo_5m/
        rel_path=     10m:topo_10m/
        rel_path= default:topo_2m/
===============================
name=LANDUSEF
        priority=1
        dest_type=categorical
        z_dim_name=land_cat
        landmask_water =   modis_30s:17            # Calculate a landmask from this field
        landmask_water = modis_lakes:17,21         # Calculate a landmask from this field
        landmask_water =  usgs_lakes:16,28         # Calculate a landmask from this field
        landmask_water =     default:16            # Calculate a landmask from this fieldz
        dominant=LU_INDEX
        interp_option =       3s:nearest_neighbor   
        interp_option =   modis_30s:nearest_neighbor
        interp_option =         30s:nearest_neighbor
        interp_option =  usgs_lakes:nearest_neighbor
        interp_option = modis_lakes:nearest_neighbor
        interp_option =          2m:four_pt
        interp_option =          5m:four_pt
        interp_option =         10m:four_pt
        interp_option =      default:four_pt
        rel_path=          3s:landuse_3s/   
        rel_path=   modis_30s:modis_landuse_20class_30s/
        rel_path=         30s:landuse_30s/
        rel_path=  usgs_lakes:landuse_30s_with_lakes/
        rel_path= modis_lakes:modis_landuse_21class_30s/
        rel_path=          2m:landuse_2m/
        rel_path=          5m:landuse_5m/
        rel_path=         10m:landuse_10m/
        rel_path=     default:landuse_2m/
===============================

SRTM index:
type = continuous
signed = yes
projection = regular_ll
dx = 0.000833333333
dy = 0.000833333333
known_x = 1.0
known_y = 1.0
known_lat = 39.999583551253
known_lon = -5.000416085531
wordsize = 4
tile_x = 6001
tile_y = 6001
tile_z = 1
missing_value=-32768
tile_bdr=0
units="meters MSL"
description="SRTM Topography height"

Corine index:
type=categorical
category_min=1
category_max=24
projection=regular_ll
dx=0.000833
dy=0.000833
known_x=1.0
known_y=1.0
known_lat=30.000000
known_lon=-10.000000
wordsize=4
tile_x=24000
tile_y=18000
tile_z=1
missing_value=-99
tile_bdr=0
units="category"
description="24-category USGS from Corine"


- I've uploaded a tar.gz file with a new write_geogrid.c and a simple makefile to make things easier, that has a README inside that explains how to compile/run. The change allows 10 times less RAM use to run the Binaries.
http://dl.dropbox.com/u/58758735/HighResgeogrid.tar.gz

Finally Change the namelist.wps and run geogrid.exe:

Code: Select all
geog_data_res = '10m','5m','2m','30s','3s','3s',


Hope this helps.
PedroCorreia
 
Posts: 5
Joined: Tue May 17, 2011 2:51 am

Re: Using CORINE land uses

Postby marmander » Mon Apr 02, 2012 5:14 pm

Some updates. gdalwarp seems to behave bad if using it on .asc files, but works fine on tiff files.

In other words, begin with transforming into WGS84:
Code: Select all
gdalwarp -s_srs g250_06.prj -t_srs EPSG:4326 g250_06.tif g250_06_WGS84.tif


After that, go on with reclassify CORINE classes to USGS classes:
Code: Select all
python reclassify.py g250_06_WGS84.tif g250_06_WGS84_USGS.tif


reclassify.py is this script:
Code: Select all
from osgeo import gdal
import numpy
import math
import sys

def reclassify(source,output):
    src_ds = gdal.Open(source)
    data = src_ds.ReadAsArray(0,0,src_ds.RasterXSize,src_ds.RasterYSize)
    driver = gdal.GetDriverByName("GTiff")
    dst_ds = driver.CreateCopy(output, src_ds, 0 , [ 'COMPRESS=LZW' ])
    for i,row in enumerate(data):
        for j,column in enumerate(row):
            data[i,j]=nclass(data[i,j])

    dst_ds.GetRasterBand(1).WriteArray(data)
    dst_ds.SetProjection(src_ds.GetProjection())
    dst_ds.SetGeoTransform(src_ds.GetGeoTransform())

    driver = None
    dst_ds = None
    src_ds = None

#This just returns the new class.
def nclass(oldclass):
    newclass=[0]*45
    newclass[1]=1
    newclass[2]=1
    newclass[3]=1
    newclass[4]=1
    newclass[5]=1
    newclass[6]=1
    newclass[7]=1
    newclass[8]=1
    newclass[9]=1
    newclass[10]=1
    newclass[11]=1
    newclass[12]=2
    newclass[13]=3
    newclass[14]=3
    newclass[15]=6
    newclass[16]=6
    newclass[17]=6
    newclass[18]=2
    newclass[19]=6
    newclass[20]=6
    newclass[21]=6
    newclass[22]=6
    newclass[23]=11
    newclass[24]=14
    newclass[25]=15
    newclass[26]=7
    newclass[27]=9
    newclass[28]=9
    newclass[29]=9
    newclass[30]=19
    newclass[31]=19
    newclass[32]=19
    newclass[33]=19
    newclass[34]=24
    newclass[35]=17
    newclass[36]=17
    newclass[37]=17
    newclass[38]=17
    newclass[39]=17
    newclass[40]=16
    newclass[41]=16
    newclass[42]=16
    newclass[43]=16
    newclass[44]=16
    return newclass[oldclass]

if len(sys.argv)==3:
      reclassify(sys.argv[1],sys.argv[2])


Now we have a reprojected, reclassified file. Now go on with file format conversion:
Code: Select all
gdal_translate -of "AAIGrid" g250_06_WGS84_USGS.tif g250_06_WGS84_USGS.asc


Now you should have a file that you can use as input to the above mentioned program, "ConvLANDUSE".

A short cut and paste howto for that:
Code: Select all
wget http://dl.dropbox.com/u/58758735/HighResgeogrid.tar.gz
tar -xzf HighResgeogrid.tar.gz
cd geogrid
make
./ConvLANDUSE /path/to/grid/g250_06_WGS84_USGS.asc


Note: File names reflect the 250m resolution. Just change 250 to 100, and you have the steps for 100 m resolution as well.

Note2: nodata definition seems to be missing, which will cause the ConvLANDUSE to give an error. Comment out the affected file-read in the fortran code, and recompile if this occurs.

Note3: missing_value/nodata is set to 0. Adjust index file accordingly.

Note4: I found yet another program to convert to geogrid format. It also creates the index file for you, and there is some instrictions as well: http://www.openwfm.org/wiki/How_to_convert_data_for_Geogrid
marmander
 
Posts: 5
Joined: Sun Mar 11, 2012 11:29 am

Re: Using CORINE land uses

Postby yaseminezber » Thu Dec 06, 2012 5:22 am

Hi Pedro,
I have been trying to use your code to create new Corine Landuse for WRF3.4.1. I did all reclassification and projection modification by using ArcGIS.When I used your code for geogrid part of the WRF, your code create the file required file (00001-05609.00001-04236). However, when I used it for geogrid.exe it complete the process but not plot anything for my domain (blank domain for both topography and landuse).
I appreciate if you have any suggestion.


PedroCorreia wrote:Hi,

first, thanks for the step by step manual. It was very helpful. I've managed to use SRTM (90m) and the Corine Land successfully in the last WRF version (3.3.1).

I just wanted to add some extra stuff i found to be needed (Under Linux):

- Convert Geotiff to asc (Both Corine and SRTM):
Code: Select all
gdal_translate -of “AAIGrid” $SRTM_FILE $File.asc


- Reclassify/Reproject Corine Land use to USGS
Install the open source GIS (grass) and do a r.reclass, using the Pineda et al. 2004 article as base.
Take in account that you need to obtain 24 Land use classes (1 to 24), and not (101 to 124).
Reproject the new raster with Lat/Lon projection into WGS84.

- Change the index and GEOGRID.TBL files:
GEOGRID.TBL
Code: Select all
===============================
name = HGT_M
        priority = 1
        dest_type = continuous
        df_dx=SLPX
        df_dy=SLPY
        smooth_option = smth-desmth_special; smooth_passes=1
        fill_missing=0.
   interp_option =      3s:four_pt
        interp_option =     30s:average_gcell(4.0)+four_pt+average_4pt
        interp_option =      2m:four_pt
        interp_option =      5m:four_pt
        interp_option =     10m:four_pt
        interp_option = default:four_pt
   rel_path=      3s:topo_3s/
        rel_path=     30s:topo_30s/
        rel_path=      2m:topo_2m/
        rel_path=      5m:topo_5m/
        rel_path=     10m:topo_10m/
        rel_path= default:topo_2m/
===============================
name = HGT_U
        output_stagger = U
        priority = 1
        dest_type = continuous   
   interp_option =      3s:four_pt
        interp_option =     30s:average_gcell(4.0)+four_pt+average_4pt
        interp_option =      2m:four_pt
        interp_option =      5m:four_pt
        interp_option =     10m:four_pt
        interp_option = default:four_pt
        smooth_option = smth-desmth_special; smooth_passes=1
        fill_missing=0.
   rel_path=      3s:topo_3s/
        rel_path=     30s:topo_30s/
        rel_path=      2m:topo_2m/
        rel_path=      5m:topo_5m/
        rel_path=     10m:topo_10m/
        rel_path= default:topo_2m/
===============================
name = HGT_V
        output_stagger = V
        priority = 1
        dest_type = continuous
   interp_option =      3s:four_pt
        interp_option =     30s:average_gcell(4.0)+four_pt+average_4pt
        interp_option =      2m:four_pt
        interp_option =      5m:four_pt
        interp_option =     10m:four_pt
        interp_option = default:four_pt
        smooth_option = smth-desmth_special; smooth_passes=1
        fill_missing=0.
   rel_path=      3s:topo_3s/
        rel_path=     30s:topo_30s/
        rel_path=      2m:topo_2m/
        rel_path=      5m:topo_5m/
        rel_path=     10m:topo_10m/
        rel_path= default:topo_2m/
===============================
name=LANDUSEF
        priority=1
        dest_type=categorical
        z_dim_name=land_cat
        landmask_water =   modis_30s:17            # Calculate a landmask from this field
        landmask_water = modis_lakes:17,21         # Calculate a landmask from this field
        landmask_water =  usgs_lakes:16,28         # Calculate a landmask from this field
        landmask_water =     default:16            # Calculate a landmask from this fieldz
        dominant=LU_INDEX
        interp_option =       3s:nearest_neighbor   
        interp_option =   modis_30s:nearest_neighbor
        interp_option =         30s:nearest_neighbor
        interp_option =  usgs_lakes:nearest_neighbor
        interp_option = modis_lakes:nearest_neighbor
        interp_option =          2m:four_pt
        interp_option =          5m:four_pt
        interp_option =         10m:four_pt
        interp_option =      default:four_pt
        rel_path=          3s:landuse_3s/   
        rel_path=   modis_30s:modis_landuse_20class_30s/
        rel_path=         30s:landuse_30s/
        rel_path=  usgs_lakes:landuse_30s_with_lakes/
        rel_path= modis_lakes:modis_landuse_21class_30s/
        rel_path=          2m:landuse_2m/
        rel_path=          5m:landuse_5m/
        rel_path=         10m:landuse_10m/
        rel_path=     default:landuse_2m/
===============================

SRTM index:
type = continuous
signed = yes
projection = regular_ll
dx = 0.000833333333
dy = 0.000833333333
known_x = 1.0
known_y = 1.0
known_lat = 39.999583551253
known_lon = -5.000416085531
wordsize = 4
tile_x = 6001
tile_y = 6001
tile_z = 1
missing_value=-32768
tile_bdr=0
units="meters MSL"
description="SRTM Topography height"

Corine index:
type=categorical
category_min=1
category_max=24
projection=regular_ll
dx=0.000833
dy=0.000833
known_x=1.0
known_y=1.0
known_lat=30.000000
known_lon=-10.000000
wordsize=4
tile_x=24000
tile_y=18000
tile_z=1
missing_value=-99
tile_bdr=0
units="category"
description="24-category USGS from Corine"


- I've uploaded a tar.gz file with a new write_geogrid.c and a simple makefile to make things easier, that has a README inside that explains how to compile/run. The change allows 10 times less RAM use to run the Binaries.
http://dl.dropbox.com/u/58758735/HighResgeogrid.tar.gz

Finally Change the namelist.wps and run geogrid.exe:

Code: Select all
geog_data_res = '10m','5m','2m','30s','3s','3s',


Hope this helps.
yaseminezber
 
Posts: 1
Joined: Thu Dec 06, 2012 4:16 am

Re: Using CORINE land uses

Postby bob.bobb » Thu Mar 21, 2013 12:35 pm

Hello,

I am experiencing the same problem like yaseminezber. I create the WPS-readable Corine-file (in my case it's 00001-38967.00001-26903), but my geo_em files are just water resp. a bunch of zero-values.

My index file is:

Code: Select all
type=categorical
category_min=1
category_max=24
projection=regular_ll
dx=0.001299
dy=0.001299
known_x=1.0
known_y=1.0
known_lat=35.094070
known_lon=-19.278683
wordsize=4
tile_x=38967
tile_y=26903
tile_z=1
missing_value=-9999
tile_bdr=0
units="category"
description="24-category USGS from Corine"


Any suggestions what could be the problem?

Could a wrong projection of my original .tif-file be the cause?

Do I need to change the Landuse.tbl in any sort?

Do I need to change/re-compile the geogrid-program itself to get it running? More specificly, are the changes inside the geogrid/src/ described by darnold in the third post only needed to obtain a WPS-readable file or also to allow geogrid.exe to read it properly?

I appreciate any help. Thanks!
bob.bobb
 
Posts: 11
Joined: Thu Mar 14, 2013 7:47 am

Next

Return to Working with Various Datasets

Who is online

Users browsing this forum: No registered users and 3 guests