170 likes | 314 Views
Destriping VIIRS visible bands (preliminary results). Karlis Mikelsons Ocean Color Group Meeting May 30, 2014. Different kinds of striping. Minor d ifferences in detector calibration and sensitivity Small amplitude No direct relation with detector number Affects all bands
E N D
Destriping VIIRS visible bands(preliminary results) KarlisMikelsons Ocean Color Group Meeting May 30, 2014
Different kinds of striping • Minor differences in detector calibration and sensitivity • Small amplitude • No direct relation with detector number • Affects all bands [M. Bouali, J. Atm. & Oc. Tech.31, 150] • Differences among detector azimuthal and zenith angles • Linear dependence on detector • Very strong near glint areas • Affects reflective bands [Q. Liu, J. Atm. & Oc. Tech. 30, 2478] • Differences in sensor mirror side (MODIS) • Alternating values for even – odd scans • Nearly same offset for all pixels within one scan • Most pronounced in MODIS blue bands A: VIIRS, nLw(672) C: MODIS, nLw(412) B: VIIRS, nLw(412)
Summary • Existing destriping algorithm still mostly works for larger amplitude saw-like striping in reflective bands • Saw-like striping in bow tie area may not be completely removed (striping might be reintroduced after remapping of coordinates) • Striping affects deletions of nLw data in 746 & 865 nm bands Things to do: • Adjust destriping parameters to optimal values • Test more examples in different conditions • Explore possibility of destriping of L1b reflectances directly • Destripe MODIS visible bands