MERIS level2#hdf!->nc ?

[slava.suslin]

Good day,

All level2 (CZCS, SeaWiFS, MODIS, VIIRS) request data have nc-format.
Why do the level2 MERIS data store in hdf-format?

thank you,
slava

[gnwiii]

Older reprocessings used HDF4 format. Check Reprocessing History to see when MERIS was last reprocessed.

[slava.suslin]

Thank You for your replay.

I'll try to explain my problem. I use the next procedure for nc - > ascii

when i work with the MERIS RR-data
/home/slava/seadas-7.3/bin/gpt.sh Write -Ssource=/media/slava/disksdc/DATA/Processing/MERIS/RR/M2004155082030.L2_RR_OC.x.hdf -PformatName=CSV -Pfile="/media/slava/disksdc/DATA/Processing/MERIS/RR/M2004155082030.L2_RR_OC.x.csv"
This works correct!
The same result i got when i work with SeaWiFS (gac, mlac), MODIS-Aqua/Terra, VIIRS. All are ok!

when i work with the MERIS FRS-data
slava@anton:/media/slava/disksdc/DATA/Processing/MERIS/FRS$ /home/slava/seadas-7.3/bin/gpt.sh Write -Ssource=/media/slava/disksdc/DATA/Processing/MERIS/FRS/M2004155083552.L2_FRS_OC.x.hdf -PformatName=CSV -Pfile="/media/slava/disksdc/DATA/Processing/MERIS/FRS/M2004155083552.L2_FRS_OC.x.csv"
This doesn't work.
I got the next report
===
org.esa.beam.framework.gpf.OperatorException: Cannot construct DataBuffer.
  at org.esa.beam.framework.gpf.internal.OperatorExecutor$GPFImagingListener.errorOccurred(OperatorExecutor.java:373)
  at com.sun.media.jai.util.SunTileScheduler.sendExceptionToListener(SunTileScheduler.java:1646)
  at com.sun.media.jai.util.SunTileScheduler.scheduleTile(SunTileScheduler.java:921)
  at javax.media.jai.OpImage.getTile(OpImage.java:1129)
  at com.sun.media.jai.util.RequestJob.compute(SunTileScheduler.java:247)
  at com.sun.media.jai.util.WorkerThread.run(SunTileScheduler.java:468)
Caused by: org.esa.beam.framework.gpf.OperatorException: Cannot construct DataBuffer.
  at org.esa.beam.gpf.operators.standard.WriteOp.computeTile(WriteOp.java:313)
  at org.esa.beam.framework.gpf.internal.OperatorImage.computeRect(OperatorImage.java:77)
  at javax.media.jai.SourcelessOpImage.computeTile(SourcelessOpImage.java:137)
  at com.sun.media.jai.util.SunTileScheduler.scheduleTile(SunTileScheduler.java:904)
  ... 3 more
Caused by: java.lang.RuntimeException: Cannot construct DataBuffer.
  at com.sun.media.jai.util.DataBufferUtils.constructDataBuffer(DataBufferUtils.java:132)
  at com.sun.media.jai.util.DataBufferUtils.createDataBufferFloat(DataBufferUtils.java:214)
  at javax.media.jai.ComponentSampleModelJAI.createDataBuffer(ComponentSampleModelJAI.java:271)
  at javax.media.jai.RecyclingTileFactory.createTile(RecyclingTileFactory.java:389)
  at javax.media.jai.PlanarImage.createWritableRaster(PlanarImage.java:1982)
  at javax.media.jai.PlanarImage.getData(PlanarImage.java:2160)
  at javax.media.jai.PlanarImage.getData(PlanarImage.java:2016)
  at javax.media.jai.RenderedOp.getData(RenderedOp.java:2266)
  at com.bc.ceres.glevel.MultiLevelImage.getData(MultiLevelImage.java:59)
  at org.esa.beam.csv.dataio.writer.CsvProductWriter.writeBandRasterData(CsvProductWriter.java:138)
  at org.esa.beam.gpf.operators.standard.WriteOp.writeTileRow(WriteOp.java:361)
  at org.esa.beam.gpf.operators.standard.WriteOp.computeTile(WriteOp.java:279)
  ... 6 more

Error: Cannot construct DataBuffer.
===

Is this memory problem? Can i edit any file (up to size memory)? What is the name of this file?

Thank You.
slava

[gnwiii]

If you are still using SeaDAS 7.3 you should upgrade to version 7.4. 

Java holds most runtime data in the "heap" area.  Unlike many applications that allocate more heap "on demand), the Java heap is limited to a maximum size that can be adjusted using the -Xmx parameter in the gpt.sh script. 

Too small heap space is a common (but not the only possible) reason for the "org.esa.beam.framework.gpf.OperatorException: Cannot construct DataBuffer." error.

I see you already posted a similar queryon the BEAM forum, where it was suggested that you increase the value of the -Xmx parameter in gpt.sh.  Have you tried this on your current system? 

On strategy for finding a suitable value is to keep doubling the memory limit until your job runs or you get a different error, e.g., try -Xmx1024M (1G), then -Xmx2048M (2G), then -Xmx4096M (4G), and so on.

[slava.suslin]

thank You.

I changed upto 4Gb, all work!

Sorry, i forgot about my question to BEAM-team ...

[gnwiii]

No need to apologize -- if it wasn't for Goggle I'd have no memory at all.  Hope ESA/Brockmann keeps the old forums online for years to come.