OceanColor Banner Image


This page contains a brief description of the various products which can be output using MSL12, with references for further information. Where possible, the algorithms are described in terms of product names (e.g., nLw_nnn, where nnn is a sensor wavelength), with links to the decription of those intermediate products.


Lw_nnn: Water-leaving radiance

The water-leaving radiance is defined as the upwelling radiance just above the sea-surface. In MSL12 product terms, it is computed as:

Lw_nnn = [(Lt_nnn - tLf_nnn) /t_oz_sen_nnn /t_oz_sol_nnn /polcor_nnn - TLg_nnn - Lr_nnn - La_nnn] /t_sen_nnn * t_oz_sol_nnn


References:
Gordon, H. R., and Wang, M. (1994), Retrieval of water-leaving radiance and aerosol optical thickness over the oceans with SeaWiFS: a preliminary algorithm. Appl. Opt. 33:443-452.


Es_nnn: Extraterrestrial solar irradiance

The solar irradiance which reaches the earth, after passage through the atmosphere, is given by:

Es_nnn = F0_nnn * t0 * mu0 * fsol

where:
t_sol_nnn = diffuse transmittance from Sun to sensor
t_oz_sol_nnn = transmittance from sun to sensor through ozone layer
mu0 = cos(solz)
fsol = earth-sun distance correction

nLw_nnn: Normalized water-leaving radiance

Normalized water-leaving radiance is defined to be the upwelling radiance just above the sea surface, in the absence of an atmosphere, and with the sun directly overhead. It is computed from Lw as:

nLw_nnn = Lw_nnn / ( t0 mu0 fsol ) * foob * brdf_nnn

where:
t0 = t_oz_sol_nnn * t_sol_nnn
mu0 = cos(solz)
fsol = earth-sun distance correction
foob = correction for out-of-band response
brdf_nnn = bi-directional reflectance correction
solz = solar zenith angle
t_sol_nnn = diffuse transmittance from Sun to sensor
t_oz_sol_nnn = transmittance from sun to sensor through ozone layer

If the outband_opt parameter is not set (outband_opt < 2), then the value of foob is 1.0. By default, the outband_opt is enabled, and foob is determined by the retieved band ratios and a bio-optical model. In that case, it attempts to correct the retrieved nLw from the full bandpass averaged value to a 10-nm square bandpass centered on the sensor nominal wavelength. This is done because most derived product algorithms are based on in situ measurements collected with narrow-band instruments, and we want the satellite retrieved radiances to be compatible with those algorithms.

The meaning of brdf_nnn will vary with the parameter setting for brdf_opt. By default, brdf_opt=7, which means brdf_nnn will correct for:
1) upwelling reflection-refraction effects at the sea-air interface
2) downwelling reflection-refraction effects at the air-see interface
3) effects of non-isotropic nature of the upwelling light field (f/Q)


Gordon, H. R., and Wang, M. (1994), Retrieval of water-leaving radiance and aerosol optical thickness over the oceans with SeaWiFS: a preliminary algorithm. Appl. Opt. 33:443-452.


Rrs_nnn: Remote sensing reflectance

Remote sensing reflectance is the standard input to many of the derived product algorithms. It is computed as

Rrs_nnn = nLw_nnn / F0_nnn,

where F0 is mean solar irradiance and nnn is wavelength. The "best" value of F0 depends on whether nLw is computed as a full bandpass value, or nominal band center value (outband_opt=2), meaning that it is corrected to a 10-nm square bandpass, centered at the sensor nominal wavelength. For standard processing, the nLw is reported as a nominal band value, and the applied F0 is therefore:

SeaWiFS Nominal Band Solar Irradiances

Wavelength (nm)mW/cm^2/um
412 171.18
443 188.76
490 193.38
510 192.56
555 183.76
670 151.22
765 123.91
865 95.965
MODIS (Aqua or Terra) Nominal Band Solar Irradiances

Wavelength (nm)mW/cm^2/um
412 171.18
443 188.76
469 203.52
488 194.18
531 185.94
551 187.00
555 183.76
645 158.74
667 152.44
678 148.14
748 127.60
859 95.728
869 94.874
1240 45.52
1640 22.99
2130 9.614
OCTS Nominal Band Solar Irradiances

Wavelength (nm)mW/cm^2/um
412 171.18
443 188.76
490 193.38
520 180.02
565 178.96
670 151.22
765 123.91
865 95.964
CZCS Nominal Band Solar Irradiances

Wavelength (nm)mW/cm^2/um
443 188.76
520 180.02
550 187.24
670 151.22
750 127.37
MOS Nominal Band Solar Irradiances

Wavelength (nm)mW/cm^2/um
408 171.08
443 188.76
485 197.73
520 180.02
570 179.59
685 146.46
750 127.37
870 94.327
OSMI Nominal Band Solar Irradiances

Wavelength (nm)mW/cm^2/um
412 171.18
443 188.76
490 193.38
555 183.76
765 123.91
865 95.964

If the nLw was not corrected to the nominal band center wavelength (i.e., outband_opt < 2) then the correct F0 is the full bandpass average. These are listed here.

K_490: Diffuse attenuation coefficient at 490 nm

SeaWiFS, MODISA,
Product info

MSl12 Source:
get_Kd.c


SST: Sea Surface Temperature

The basis for the SST algorithm:




MSl12 Source:
sst.c

References:
SST ATBD


chlor_a: sensor default chlorophyll algorithm

Each sensor is assigned a default chlorophyll algorithm, which can be output as product chlor_a. The default algorithm varies by sensor due to limitations on the available spectral bands. The default chlorophyll is also used for any intermediate calculations which require chlorophyll (e.g., f/Q bi-directional reflectance corrections). The default algorithms by sensor are:

SeaWiFSchl_oc4
MODIS/Aquachl_oc3
MODIS/Terrachl_oc3
OCTSchl_oc4
POLDERchl_oc4
CZCSchl_oc2
MOSchl_oc4
OSMIchl_oc2


chl_oc2: Chlorophyll-a concentration, OC2 algorithm

The OC2 algorithm is simple band ratio algorithm developed by O'Reilly et al. as a product of the SeaWiFS Bio-optical Algorithm Mini Workshop (SeaBAM). The algorithm inputs are the retrieved remote sensing reflectances, Rrs_nnn.

R = Log10(Rrs_490/Rrs_555) or Log10(Rrs_490/Rrs_565)
chl_oc2 = 10.0^(a[0]+a[1]*R+a[2]+R^2+a[3]*R^3) + a[4]

Where the a[] coefficients were derived by fitting to the SeaBAM in situ dataset. Fits have been done for both the 490 to 555 combination and the 490 to 565 combination. MSl12 will use the coefficients which correspond most closely to the sensor nominal band center wavelengths. The fit coefficients for the OC2 algorithm are:

a = {0.3164,-2.1320,0.6303, 0.0040,-0.0708}, for OCTS & POLDER
a = {0.2974,-2.2429,0.8358,-0.0077,-0.0929}, for all other sensors

Algorithm Failure Conditions:
chl_oc2 will be set to -1.0 and the CHLFAIL flag will be set if
Rrs_490 < 0
Rrs_555 or Rrs_565 < 0
R > 10.0

Algorithm Warning Conditions:
The CHLRANGE flag will be set if
chl_oc2 <= 0.0
chl_oc2 > 100.0

MSl12 Source:
Functions get_chl_oc2v2() & get_chl_oc2_octs() in get_chl.c

References:
  • O'Reilly, J. E., S. Maritorena, B. G. Mitchell, D. A. Siegel, K. L. Carder, S. A. Garver, M. Kahru, and C. McClain (1998),
    Ocean color chlorophyll algorithms for SeaWiFS, J. Geophys. Res., 103(C11), 24,937-24,954.


    • chl_oc3: Chlorophyll-a concentration, OC3 algorithm




    Algorithm Failure Conditions:
    chl_oc3 will be set to -1.0 and the CHLFAIL flag will be set if
    Rrs_555 < 0
    Rrs_443 or Rrs_490 < 0
    R > 10.0

    Algorithm Warning Conditions:
    The CHLRANGE flag will be set if
    chl_oc3 <= 0.0
    chl_oc3 > 100.0

    MSl12 Source:
    Function get_chl_oc3_modis() in     get_chl.c

    References:
  • O'Reilly et al. 2000
  • Evaluation of OC3 algorithm
  • Validation Results


    chl_oc4: Chlorophyll-a concentration, OC4 algorithm

    OC4 is a modified cubic polynomial, four band (443,490,510,555), maximum band ratio alogrithm. Whichever ratio is greatest (Rrs443/Rrs555, or Rrs490/Rrs555, or Rrs510/Rrs555) determines the band ratio.


    Algorithm Failure Conditions:
    chl_oc4 will be set to -1.0 and the CHLFAIL flag will be set if
    Rrs_570 < 0
    Rrs_443 or Rrs_510 < 0
    R > 10.0

    Algorithm Warning Conditions:
    The CHLRANGE flag will be set if
    chl_oc4 <= 0.0
    chl_oc4 > 100.0

    MSl12 Source:
    Functions get_chl_oc4v4, get_chl_oc4v4_mos(), & get_chl_oc4v4_octs(), in     get_chl.c

    References:
  • O'Reilly, J. E., S. Maritorena, B. G. Mitchell, D. A. Siegel, K. L. Carder, S. A. Garver, M. Kahru, and C. McClain (1998),
    Ocean color chlorophyll algorithms for SeaWiFS, J. Geophys. Res., 103(C11), 24,937-24,954.
  • O'Reilly et al. 2000
  • evaluation of OC4 algorithm


    chl_octs: OCTS Chlorophyll a concentration (OCTS-C algorithm)




    MSl12 Source:
    Function get_chl_octsc in     get_chl.c


    Normalized Difference Pigment algorithm Chlorophyll Concentration

    All Sensors

    chl_clark: Chlorophyll-a concentration, Clark Empirical Algorithms

    This is a three band empirical chlorophyll algorithm



    MSl12 Source:
    Functions get_chl_clark_modis, get_chl_clark_seawifs in get_chl.c

    Other Clark empirical Algorithms:

  • poc_clark, particulate organic carbon

    MSl12 Source:
    Function get_poc.c

  • tsm_clark, total suspended matter

    MSl12 Source:
    get_tsm.c


    • References:
    Clark ATBD

    Garver-Siegel-Maritorena-2001 Semi-Analytical Bio-Optical Model
  • chl_gsm01, chlorophyll concentration

  • a_nnn_gsm01, total absorption at sensor wavelength nnn

  • bb_nnn_gsm01, total backscatter at sensor wavelength nnn

  • bbp_nnn_gsm01, particulate backscatter at sensor wavelength nnn

  • aph_nnn_gsm01, absorption due to phytoplankton at sensor wavelength nnn

  • adg_nnn_gsm01, absorption due to gelbstof and detrital material at sensor wavelength nnn

    • MSl12 Source:
    gsm01.c and amoeba.c

    References:
    Maritorena_et_al 2002

    Spectral Kd Algorithm

     MSl12 Source:
    get_Kd.c KDtree.c

    References:
    ZPLee_2004

    Quasi-Analytical Algorithm of Z.P.Lee
  • a_nnn_qaa QAA (Quasi-Analytical Algorithm) model (Z.P.Lee) total absorption at sensor wavelength nnn
  • bb_nnn_qaa QAA (Quasi-Analytical Algorithm) model (Z.P.Lee) total backscatter at sensor wavelength nnn
  • bbp_nnn_qaa QAA (Quasi-Analytical Algorithm) model (Z.P.Lee) particulate backscatter at sensor wavelength nnn
  • aph_nnn_qaa QAA (Quasi-Analytical Algorithm) model (Z.P.Lee) absorption due to phytoplankton at sensor wavelength nnn
  • adg_nnn_qaa QAA (Quasi-Analytical Algorithm) model (Z.P.Lee) absorption due to gelbstof and detrital material at sensor wavelength nnn


    • MSl12 Source:
    get_qaa.c, qaa.c

    References:
    Lee,Z.P., K.P.Du, and R. Arnone, "A Model for the diffuse attenuation coefficient of downwelling irradiance."
    J.Geophys.Res.,accepted, 2004b.

    Carder IPAR & ARP
  • ipar, instantaneous photosynthetically available radiation
  • arp, instantaneous absorbed radiation by phytoplankton


    • MSl12 Source:
    ipar_arp.c

    References:
    Carder Ipar and Arp ATBD

    Gordon & Balch Calcite
  • calcite_2b, calcite concentration, 2-band algorithm
  • calcite_3b, calcite concentration, 3-band algorithm

  • calcite, calcite concentration, merged algorithm (DEFAULT FOR ALL SENSORS)

    • MSl12 Source:
    calcite.c

    References:
    Calcite ATBD

    Chlorophyll Fluorescence and Supporting Products
  • flh, fluorescence line height
  • cfe, chlorophyll fluorescence efficiency
  • Support for 678nm channel (e.g., nLw_678, tau_678, etc.)

    • MSl12 Source:
    fluorescence.c

    References:
    Fluorescence ATBD

    This is a SeaWiFS-only product.

    MSl12 Source:
    Function in get_par.c

    Level 2 processing flags

    The l2_flags output has the same definition as used in the SeaWiFS L2 data. This program allows certain flags to be used as masks. Processing is bypassed for masked pixels and the output value is set to zero.



    Land Products:

    rhos_nnn Surface reflectance
    ndvi Normalized Difference Vegetation Index



    Observed Radiances and Reflectances:

    Lt_nnn: Calibrated TOA radiance mw/cm^2/um/sr
    rhot_nnn: Top of atmosphere reflectance


    t_sol_nnn: diffuse transmittance, Sun to ground

    This product gives the transmittance of diffuse skylight along the path from the Sun to the ground pixel. It accounts for the probability of scattering losses due to Rayleigh and aerosol. It is therefore dependent on the aerosol type, as determined by the aerosol models which were selected within the atmospheric correction process. The Rayleigh-aerosol diffuse transmittance is parameterized as a series of exponential functions defined at various zenith angles for each aerosol model, at each sensor wavelength. The exponentials are of the form a*exp(-b*Taua_nnn), where Taua_nnn is the aerosol optical thickness at wavelength nnn. The a and b coefficients are stored in the aerosol model look-up tables.

    t_sen_nnn: diffuse transmittance, ground to Sun

    This product gives the transmittance of a diffuse light source along the path from the ground pixel to the Sun. It accounts for the probability of scattering losses due to Rayleigh and aerosol. It is therefore dependent on the aerosol type, as determined by the aerosol models which were selected within the atmospheric correction process. The Rayleigh-aerosol diffuse transmittance is parameterized as a series of exponential functions defined at various zenith angles for each aerosol model, at each sensor wavelength. The exponentials are of the form a*exp(-b*Taua_nnn), where Taua_nnn is the aerosol optical thickness at wavelength nnn. The a and b coefficients are stored in the aerosol model look-up tables.

    t_oz_sol_nnn: ozone transmittance, Sun to ground

    This product gives the transmittance of solar radiation through the ozone layer, along the path from the Sun to the ground pixel. The calculation accounts for the actual ozone concentration and the slant path through the ozone layer as defined by the solar zenith angle, product solz. For each sensor band, it is given by:

    t_oz_sol_nnn = exp(-(o3 * k_oz_nnn)/cos(solz))


    where:
    o3 = ozone concentration, as provided by the ancillary input meterological files
    k_oz_nnn = is the ozone absorption cofficient for sensor band nnn, as provided in the msl12_sensor_info.dat files


    t_oz_sen_nnn: ozone transmittance, ground to Sun

     This product gives the transmittance of solar radiation through the ozone layer, along the path from the ground pixel to the Sun. The calculation accounts for the actual ozone concentration and the slant path through the ozone layer as defined by the sensor zenith angle, product senz. For each sensor band, it is given by:

    t_oz_sen_nnn = exp(-(o3 * k_oz_nnn)/cos(senz))


    where:
    o3 = ozone concentration, as provided by the ancillary input meterological files
    k_oz_nnn = is the ozone absorption cofficient for sensor band nnn, as provided in the msl12_sensor_info.dat files


    Atmospheric Correction and Other Intermediate Products:

    t_o2_nnn Total oxygen transmittance
    glint_coeff : Glint radiance normalized by solar irradiance (ocean only)
    aerindex : Aerosol index (for identification of absorbing aerosols) (ocean only)
    cloud_albedo: Reflectance used for cloud/ice thresholding (historical name)
    aer_model
    aer_model_min Minimum bounding aerosol model # (ocean only)
    aer_model_max Maximum bounding aerosol model # (ocean only)
    aer_model_ratio Model mixing ratio (ocean only)
    aer_num_iter Number of aerosol iterations, NIR correction (ocean only)
    epsilon Retrieved epsilon used for model selection at 765 and 865 nm (float format)




    eps_78: Alternate name for epsilon (scaled to byte)








    angstrom_nnn: Aerosol angstrom coefficient






    taua_nnn: Aerosol optical depth




    Lr_nnn: Rayleigh radiance mw/cm^2/um/sr
    La_nnn: Aerosol radiance (ocean only) mw/cm^2/um/sr
    TLg_nnn: Top-of-atmosphere(TOA) glint radiance (ocean only) mw/cm^2/um/sr
    tLf_nnn: Foam (white-cap) radiance (ocean only) mw/cm^2/um/sr
    brdf_nnn: Bi-directional reflectance correction factor. (Morel, et al.) (ocean only)
    fsol: Solar distance correction factor per scan


    Input Ancillary Products:

     ozone: Ozone concentration (from input ancillary data)
    windspeed: Magnitude of wind (m/s)
    windangle: Wind direction (deg) N=0, E=90
    zwind: Zonal wind speed (m/s)
    mwind: Meridional wind speed (m/s)
    water_vapor: Precipitable water concentration g/cm^2
    pressure: Barometric Pressure mb
    humidity: Relative Humidity %
    sstref: Sea Surface Temperature (interpolated from climatology to pixel location) deg C

    Geometry:

    solz: Solar zenith angle

     The angle between the local zenith and the line of sight to the sun




    sola: Solar azimuth angle deg
    senz: Sensor zenith angle deg
    sena: Sensor azimuth angle deg
    height: Terrain height m