MODIS-CE Description

Measurements

The MODerate-resolution Imaging Spectroradiometer (MODIS) on the NASA EOS platforms, Terra and Aqua, provide spectral radiances measurements in 32 solar and infrared bands since February 2000 and July 2002, respectively. The common resolution of 1 km is used for the 5 bands (0.64, 3.8, 1.6 or 2.1, 10.8, and 12.0 μm) employed in the Clouds and the Earth's Energy System (CERES) Edition 2 cloud detection and retrieval. Prior to processing, the MODIS data sampled every other scan line and every 4th pixel, resulting in a nominal resolution of ≈2.8 km. Clouds are detected using a cascading threshold method applying sets of tests based on expected clear-sky radiances or brightness temperature differences and variability. Retrievals of cloud properties use subsets of the five channels and assume a plane-parallel single layer cloud completely covers each MODIS cloudy pixel. The sampled MODIS pixel results are averaged to obtain mean values for the clear portion of 1° x 1° grid and the clouds at up to two layers.

Local Observation Time and Length of Data Record

Terra, 10:30 AM and 10:30 PM ; 2001–2009
Aqua, 1:30 AM and 1:30 PM ; 2003–2008

Spatial Resolution

2 km (sampled), 4 km (sampled), 32-km clear sky radiance estimation

Cloud Detection

Thresholds based on standard deviation of expected clear radiances + expected clear radiance estimate at 0.64, 10.8 μm and for brightness temperature difference (BTD) between 3.8 and 10.8 μm during day; same during night and twilight (solar zenith angle = 82–89°) except no 0.64 μm data used and add BTD between 10.8 and 12 μm. Thin cirrus check used day and night as variable threshold compared to BTD(10.8–12). Multiple special checks are applied, especially over polar regions.

Retrieval Methodology

Daytime: 0.64, 3.8, and 10.8-μm radiances used simultaneously to estimate cloud phase, COD, CT, and CRE by matching to model calculations employing cloud reflectance LUTs and emittance parameterizations, and surface and clear-sky radiances. CEM computed from COD. CT → CZ using GEOS-4 temperature profile altered by simple 7.1 K/km lapse rate in the boundary layer; CZ → CP; COD + CRE → CWP.
If no solution is found, a bispectral method is used to estimate the initial variables; other wise, the pixel is declared a no-retrieval pixel and no values are reported. Over snow and ice surfaces, either the 1.6-μm (Terra) or 2.1-μm (Aqua) reflectance replaces the 0.64-μm reflectance. Aqua snow-ice retrieval of COD underestimated due to code bug and insensitivity for ice clouds over snow at 2.1 μm.
Nighttime: 10.8-μm radiance, BTD(3.8–10.8), and BTD(10.8–12) used simultaneously to estimate cloud phase, COD, CT, and CRE by matching to model calculations employing cloud emittance parameterizations. If COD > 8, default COD values of 8, 16, or 20 are assigned depending on CT and other factors. CEM computed from COD. CT → CZ using GEOS-4 temperature profile altered by simple 7.1 K/km lapse rate in the boundary layer; CZ → CP; COD + CRE → CWP.

Ancillary Input

  • GEOS-4 vertical profiles of temperature/humidity (1° spatial, 6 h temporal, 58 layers)
  • GEOS-4 surface skin temperatures (1° spatial, 3 h temporal resolutions)
  • NCEP SMOBA ozone total optical depth (2.5° resolution, daily)
  • IGBP surface type (19 types including coast, 10' resolution)
  • SSM/I snow and ice cover (daily, 25-km)
  • NESDIS Multisensor snow and ice cover (daily, interpolated to 10')
  • CERES MODIS surface emissivity (3.8, 10.8, 12 μm, 10', monthly)
  • CERES MODIS clear-sky albedo (0.64, 1.6/2.1 μm, 10', monthly, updated bi-daily)
  • Clear-sky/surface directional/bidirectional reflectance anisotropic models (surface type)
  • USGS GTOPO30 land elevation & water percentage (1-km resolution averaged to 10')
  • 4-layer adding-doubling radiative transfer parameterization (0.64 μm)
  • multi-layer radiative transfer with layer absorption/emission from correlated-k method
  • cloud emittance parameterizations (3.8, 10.8, and 12 μm)
  • water: spherical droplets, 7 sizes; ice hexagonal column distributions, 9 sizes
  • cloud reflectance look up tables (0.64 and 3.8 μm)
  • water: spherical droplets, 7 sizes; ice hexagonal column distributions, 9 sizes
  • References

  • Minnis, P., Q. Z. Trepte, S. Sun-Mack, Y. Chen, D. R. Doelling, D. F. Young, D. A. Spangenberg, W. F. Miller, B. A. Wielicki, R. R. Brown, S. C. Gibson, and E. B. Geier, 2008a: Cloud detection in non-polar regions for CERES using TRMM VIRS and Terra and Aqua MODIS data. IEEE Trans. Geosci. Remote Sens., 46, 3857-3884.
  • Minnis, P., S. Sun-Mack, D. F. Young, P. W. Heck, D. P. Garber, Y. Chen, D. A. Spangenberg, R. F. Arduini, Q. Z. Trepte, W. L. Smith, Jr., J. K. Ayers, S. C. Gibson, W. F. Miller, V. Chakrapani, Y. Takano, K.-N. Liou, Y. Xie, and P. Yang, 2011: CERES Edition-2 cloud property retrievals using TRMM VIRS and Terra and Aqua MODIS data, Part I: Algorithms. IEEE Trans. Geosci. Remote Sens., 49, 11, 4374-4400.
  • Minnis, P., S. Sun-Mack, Y. Chen, M. M. Khaiyer, Y. Yi, J. K. Ayers, R. R. Brown, X. Dong, S. C. Gibson, P. W. Heck, B. Lin, M. L. Nordeen, L. Nguyen, R. Palikonda, W. L. Smith, Jr., D. A. Spangenberg, Q. Z. Trepte, and B. Xi, 2011: CERES Edition-2 cloud property retrievals using TRMM VIRS and Terra and Aqua MODIS data, Part II: Examples of average results and comparisons with other data. IEEE Trans. Geosci. Remote Sens., 49, 11, 4401-4430.
  • Trepte, Q., P. Minnis, and R. F. Arduini, 2002: Daytime and nighttime polar cloud and snow identification using MODIS data. Proc. SPIE 3rd Intl. Asia-Pacific Environ. Remote Sensing Symp. 2002: Remote Sens. of Atmosphere, Ocean, Environment, and Space, Hangzhou, China, October 23-27, Vol. 4891, 449-459.
  • CERES Data Website

    http://ceres.larc.nasa.gov