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

MODIS landuse data

epotter1

New member
Hello,

I am trying to update the landuse input data in WPS to include a better representation of the glaciers in South America (by adjusting them based on the Randolph Glacier Inventory (RGI)). I have previously done this using the USGS input data from USGS EROS LandCover GLCCDB V2p0 data files as a Geotif, modifying it using the RGI using Arc, converting the geotiff to binary input files, running WPS and adjusting the soil parameters to match the land surface for glaciers. I am now trying to use the default modis_landuse_20class_30s_with_lakes instead of USGS. However I have run into the following problems:

1. I cannot find where to download the MODIS landuse data, or even an exact reference for it. I believe it is the Modified IGBP MODIS 20-category vegetation (land-use) data (ftp) from Noah-Multiparameterization Land Surface Model (Noah-MP® LSM) | Research Applications Laboratory , but the link to this doesn't appear to lead anywhere. Does anyone know where the MODIS landuse data used as the WPS default can be downloaded from in Geotiff (or similar) form?

2. Instead of downloading the data, I could in principle convert the static data binary files from MODIS to a geotiff or similar. I was previously using a java script to translate from PNG to the binary files needed for the static data for geogrid, which had both a Read and Write class which could also convert the other way, but I cannot quite get the geographic referencing correct. This was written by someone called Danny Brown from Edmonton, Canada. I found this on a help topic on the old WRF forum, but the link is now broken. I now only have the compiled programmes, and not the original java scripts, and therefore I can't edit the scripts to try to fix the geographic referencing. I can see there is now a python script called 'convert_geotiff' that might be able to write the binary files, but does anyone know of a program (either the java one, convert_geotiff, gdal_translate or another) which can convert the static geographic data into a Geotiff or similar?

3. If anyone knows an up-to-date landcover classification for South America, I would use this instead as I understand both the UGSG and Modis datasets are now quite old.

Many thanks for help with any of these questions!
 
Hi,

I can try to address question 1.
1. I cannot find where to download the MODIS landuse data, or even an exact reference for it. I believe it is the Modified IGBP MODIS 20-category vegetation (land-use) data (ftp) from Noah-Multiparameterization Land Surface Model (Noah-MP® LSM) | Research Applications Laboratory , but the link to this doesn't appear to lead anywhere. Does anyone know where the MODIS landuse data used as the WPS default can be downloaded from in Geotiff (or similar) form?

I believe the reference for the MODIS data is mentioned in this FAQ. Regarding the ftp site, I also tried to ftp the data to my system and it seems to make contact, but nothing ever happens. Is this what happens for you, as well? I would suggest contacting someone in NCAR/RAL to ask if they have any thoughts. At the bottom of the page you linked are two contacts you can try.
 
Thanks very much for this. Yes, all the ftp links seem to just freeze for me. I will see if I can email someone about this.

If anyone else needs to do this, I've got somewhere with the USGS data from the link in my original post. I exported the new adjusted landuse as a tif from Arcmap, and translated it to the binary file using "gdal_translate -of ENVI My_new_tif.tif binary_data_file.bil", then renamed the binary_data_file.bil as 00001-06960.00001-09960 (the data in this case is 6960 by 9960 pixels big, and I just have one binary file). I edited the index file to include the new size and the reference lat and lon (from the tif). This has almost worked, except the landuse data is upside-down. I have no idea why, but will update if I work out how to flip it!
 
Top