I want to extract chlorophyll concentration information from MODISA and converted to .csv format so I can conduct further analysis using R.
I download the data file from MODISA and the format appears to be *.main. I tried to load .main files into R but failed.
So I tried to use SeaDAS software can convert the *.main file to .csv. I first load the .main file into SeaDAS (File/ Import Raster Data/ SeaDAS L3 (CZCS, OCTS, SeaWiFS, MODIS, VIIRS, etc,)). I then tried to convert it into .csv format by File/ Export Raster Data/ CSV
But it just generated an empty CSV file.
I tried to look up to the tutorial for SeaDAS 7, but I can still not solve the problem. Could any one help?
As long as I can convert .main into to a R processable format, I can use the data for further analysis.
P.S. I run SeaDAS on Linux environment (ubuntu).
hi,
you are using a binned product from
http://oceandata.sci.gsfc.nasa.gov/MODISA/Binned
with SeaDas, do not import it as a raster. Open it as a product
File -> Open Product
After that choose your band that you want and export it as CSV
Double click on band -> File -> Export Raster Data -> CSV
It will take several minutes to export the data. Huge files.
At the bottom-right of Seadas screen, you can see the export processing. Be patient!
cheers, Kostas
Hi Kostas,
Thank you for your quick reply.
I tried to follow your suggestion; however, it appears that it takes a long time to complete the export. After wait for more than an hour, and it still appeared as "writing project..."
I checked the property of the exported .csv file and found the size of the file does not increase with time (size < 15kb). I think it is not normal and might have nothing to with the computing power of my computer.
Since you have more experience using this software, so, if it OK for you, I would like to sent you an e-mail. In that e-mail I will include a series of screen shots (about how I use this software). I hope that with your help I would find out what step did I miss or wrong.
Thanks,
Hao-Wei
hi Hao-Wei,
There is nothing wrong with the way your using SeadDas. The information you want to extract produce very big files and require a lot of time to be extracted. If you check the CPU's usage, while you extract, you will see that it is 100% (open a console and write the command "top", hit ender).
A few tips that might help you with the procedure.
1. If you do not need the global data, choose the subregion of your interest.
After the File -> Export Raster Data -> CSV in the new box that will open on the upper right corner you will find a button named "Subset". From this one you can reduce the size of the area and to choose the bands that will be extracted.
On the new box named "Specify Product Subset" I will strongly suggest to use the "Geo Coordinates" tab, in order to crop the extracted area. Always write down the Coordinates that you use and use the same subset for all of your extractions.
2. Choose only the band(s) of your interest.
In the same box named "Specify Product Subset", on the second tab named "Band subset" choose only the band(s) of your interest. i.e. "weights", "clor_a_sum", "clor_a_mean"
For a small area like Mediterranean and only for the 3 above mentioned bands, the procedure required about 30 minutes on my laptop
3. You may check your file with gdalinfo utility.
In a console write the command:
gdalinfo /where/ever/you/saved/your/file/file.name
All the above if you want to export the data in a CSV file.
My suggestion is to export the data in a GeoTiff or even a NetCDF file,which then you can easily import into R.
Another option. (I never tried it!!!)
The binned Modis files are HDF4 files. Therefore, you can convert them to HDF5 files with the utility h4toh5(http://www.hdfgroup.org/h4toh5/) and then import them into R with the package “rhdf5”
Feel free to send me an email (pm) if you have any further queries
cheers, Kostas
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R