[gdal-dev] HDF4 reading MISR Level 3 with incorrect dimension

[Lucena, Ivan]

Hi there,

I don't know if that is a case for a ticket or if I missed the latest discussion on HDF dimensions.

I've got this HDF4 file, a MISR Level 3 product [1]:

---
Subdatasets:
  SUBDATASET_1_NAME=HDF4_EOS:EOS_GRID:"MISR_AM1_CGAS_APR_2000_F15_0031.hdf":Aero
solParameterAverage:Absorbing optical depth
  SUBDATASET_1_DESC=[360x720x9x4] Absorbing optical depth AerosolParameterAverag
e (32-bit floating-point)
  SUBDATASET_2_NAME=HDF4_EOS:EOS_GRID:"MISR_AM1_CGAS_APR_2000_F15_0031.hdf":Aero
solParameterAverage:Algorithm type count
  SUBDATASET_2_DESC=[360x720x10x2] Algorithm type count AerosolParameterAverage
(32-bit integer)
  SUBDATASET_3_NAME=HDF4_EOS:EOS_GRID:"MISR_AM1_CGAS_APR_2000_F15_0031.hdf":Aero
solParameterAverage:Angstrom exponent
  SUBDATASET_3_DESC=[360x720x9] Angstrom exponent AerosolParameterAverage (32-bi
t floating-point)
  SUBDATASET_4_NAME=HDF4_EOS:EOS_GRID:"MISR_AM1_CGAS_APR_2000_F15_0031.hdf":Aero
solParameterAverage:Average fill flag
  SUBDATASET_4_DESC=[360x720] Average fill flag AerosolParameterAverage (8-bit i
nteger)
  SUBDATASET_5_NAME=HDF4_EOS:EOS_GRID:"MISR_AM1_CGAS_APR_2000_F15_0031.hdf":Aero
solParameterAverage:Optical depth average
  SUBDATASET_5_DESC=[360x720x9x4x6] Optical depth average AerosolParameterAverag
e (32-bit floating-point)
  SUBDATASET_6_NAME=HDF4_EOS:EOS_GRID:"MISR_AM1_CGAS_APR_2000_F15_0031.hdf":Aero
solParameterAverage:Optical depth average count
  SUBDATASET_6_DESC=[360x720x9x4x6] Optical depth average count AerosolParameter
Average (32-bit integer)
  SUBDATASET_7_NAME=HDF4_EOS:EOS_GRID:"MISR_AM1_CGAS_APR_2000_F15_0031.hdf":Aero
solParameterAverage:Optical depth standard deviation
  SUBDATASET_7_DESC=[360x720x9x4x6] Optical depth standard deviation AerosolPara
meterAverage (32-bit floating-point)
  SUBDATASET_8_NAME=HDF4_EOS:EOS_GRID:"MISR_AM1_CGAS_APR_2000_F15_0031.hdf":Aero
solParameterAverage:Single-scatter albedo
  SUBDATASET_8_DESC=[360x720x9x4] Single-scatter albedo AerosolParameterAverage
(32-bit floating-point)
  SUBDATASET_9_NAME=HDF4_EOS:EOS_GRID:"MISR_AM1_CGAS_APR_2000_F15_0031.hdf":Aero
solParameterAverage:Single-scatter albedo count
  SUBDATASET_9_DESC=[360x720x9x4] Single-scatter albedo count AerosolParameterAv
erage (32-bit integer)
---

If I gdal_translate (with -sds) the first dataset [360x720x9x4] to GTIFF I've got a 9 bands of 720x360. That sounds 
good but doesn't looks good unfortunately. The images looks like a sliced book as you would see it by the side, a 
typical effect of dimension order error.

Any help with that?

Regards,

Ivan

---
[1] http://eosweb.larc.nasa.gov/PRODOCS/misr/DPS/DPS_v50_RevR.pdf