Dear Researchers,
I need to download LAI data set and transform it to the WGS84 (EPSG:4326).
I found following resource:
http://glcf.umd.edu/data/lai/
As I can see in the description, following information about projection is provided only:
'It is provided in an Integerized Sinusoidal (ISIN) projection at the resolution of 1 km for each of the 289 land tiles (representing 1200 x 1200 pixels on the Earth) on the globe.'
What is an appropriate EPSG code for this projection?
How can I transform this data set? Could you suggest any tool or example script which will allow me to transform and extract needed information?
If this might help me - on my own I wrote a script to extract WGS84 coordinates (I guess that they're in WGS) coordinates from the metadata file (.hdf.xml) attached to the main data file (.hdf).
http://glcf.umd.edu/data/lai/
Dear Mateusz,
It is not an easy task which can be solved via a simple selection of the right and geometrically and geodetic fitting EPSG number for any defined coordinate system.
We've solved this task by following the description and python script from Ryan J. Dillon published on the internet under: http://skemman.is/stream/get/1946/16233/38046/1/final_processingwithpython_dillon.pdf
or the R tutorial presented under: http://menugget.blogspot.de/2012/04/working-with-globcolour-data.html from R blogger "marc in the box".
I wish you good luck and success with your interesting task.
Dear Mateusz,
according to the QGIS software, there are two global sinusoidal projections that have EPSG codes:
Sphere sinusoidal: 53008
World sinusoidal: 54008
for NASA-related products, I suggest to use the first one (the second one is the same for the WGS84 ellipsoid):
http://modis-land.gsfc.nasa.gov/GCTP.html
note that according to QGIS, the radius of the sphere is shorter by few meters than in the NASA metadata. I guess it causes almost no error if the metadata is good from all other aspects.
good luck, Gábor
The Integerized Sinusoidal data at the UMD link you refer to says its derived from MOD09A1. Its very likely that its the same LAI data (or similar) as available from the USGS Land Processes DAAC, see https://lpdaac.usgs.gov/products/modis_products_table/mod15a2.
Worth checking a few files to make sure as the UMD group may have modified their algorithm from that used to produce the LPDAAC version. I understand that is what BNU have done and the UMD and BNU groups seem to be coordinating their work.
In any case, its likely that the MODIS Reprojection Tool (https://lpdaac.usgs.gov/tools/modis_reprojection_tool) will be able to read the HDF-EOS files from the UMD group. In particular the 'resample' tool should be able to convert the sinusoidal data to WGS84 (with RESAMPLING_TYPE = NN; OUTPUT_PROJECTION_TYPE = GEO). The 'mosiac' tool may also be able to join/mosaic the hXXvYY files together for you too - if so, you need to do the mosaic step before the resample.
I haven't tested the MRT software for the UMD LAI data you link to, only for the LPDAAC data. So I hope I haven't sent you down a rabbit hole with this info.
All the best, Matt.
Here are a couple of references for the use of the Integerized Sinusoidal grid for MODIS products:
Masuoka et al (1998) Key characteristics of MODIS data products, http://dx.doi.org/10.1109/36.701081 -> section C1.
Rossow et al (1984) Selection of a map grid for data analysis and archival, http://dx.doi.org/10.1175/1520-0450(1984)0232.0.CO;2
http://modis-land.gsfc.nasa.gov/MODLAND_grid.html
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing