CZCS Atmospheric Correction
CZCS Atmospheric CorrectionLike all spaceborne radiometers, CZCS measures the spectral distribution of radiance exiting the top of the atmosphere, but for ocean color applications we require the spectral distribution of radiance upwelling from below the ocean surface. The atmospheric correction process removes the radiance contributions associated with scattering from the atmosphere (i.e., aerosols and air molecules) and surface contributions such as specular reflection and white-caps, and adjusts for attenuation of the retrieved water-leaving radiance by the atmosphere. Wherever possible, the process used for CZCS is identical to that used by the OBPG to process SeaWIFS and MODIS data (e.g., MODIS/Aqua Reprocessing 1.1 & SeaWiFS Reprocessing 5.1). The relative spectral response of each CZCS band was used to generate the Rayleigh and aerosol tables and other band-pass-specific quantities (e.g., solar irradiances, ozone absorption coefficients, out-of-band response). The only difference in the atmospheric correction process of CZCS relative to modern sensors is in the procedure and assumptions used to determine aerosol contributions.
Contributions from Aerosols
The OBPG uses the method described in Gordon & Wang (1994) to determine the aerosol type and aerosol radiance contributions for MODIS & SeaWiFS, but this method relies on the existence of a pair of near infrared (NIR) bands where the water-leaving radiance contribution is very small and predictable. An iterative scheme (Stumpf et al., 2003) is employed to predict the water-leaving radiance contribution from the two NIR bands, and thereby retrieve the aerosol radiance at two wavelengths. The retrieved spectral difference in the NIR bands is used to select the most appropriate aerosol model type, and this aerosol model, in combination with the retrieved aerosol radiance in the NIR, is then used to assess the aerosol contributions in the shorter visible bands. Unfortunately, CZCS does not have a pair of spectral channels in the NIR, and the longest functional wavelength at 750 nm was not designed for ocean observations. The longest available wavelength for CZCS ocean observations is 670 nm, which has a small but not insignificant water-leaving radiance contribution in oligotrophic and mesotrophic waters (larger and less predictable in eutrophic and turbid waters). To determine the water-leaving radiance contribution at 670 nm (and thereby the aerosol contribution at that longest wavelength), the OBPG developed an iteration scheme very similar to Stumpf et al. (2003). This was used in combination with an assumption about the aerosol type to determine the aerosol contribution in the shorter visible wavelengths.
Determination of Water-Leaving Radiance at 670 nm
As shown in the attached code snippet, a procedure was developed to estimate water-leaving reflectance at 670 from reflectance at 550. The algorithm uses a linear relationship between particulate backscattering and Rrs_550 described in Carder et. al.(1999). The spectral scattering function described by Gould et. al.(1999) is used to estimate backscattering at 670 nm from backscattering at 550. The Gordon et. al.(1988) radiance model formulation is then used to estimate reflectance at 670 from backscattering and absorption. The absorption terms account for water (computed a priori for the CZCS spectral bands), particulate matter (estimated from chlorophyll via Bricaud et. al, 1998), and gelbstof/detritus estimated via (Stumpf et. al, 2003). Thanks go to R.W.Gould, R.P.Stumpf, and R.A.Arnone for helpful discussion during the development of this technique.
Assumption on Aerosol Type
For typical maritime aerosols (e.g., non-absorbing or weekly absorbing, sea-salt and water) the aerosol type (i.e., the aerosol size distribution) is needed to translate aerosol radiance measured at one wavelength to another wavelength. However, the determination of aerosol type requires at least two wavelengths. Rather than attempting to estimate the water-leaving radiance at 550 nm, which would require assumptions about a critical quantity we wish to "measure" (assumptions that would be prone to error in all but the clearest ocean waters) the OBPG chose to make an assumption that the aerosol type was fixed to that of a maritime aerosol at 99% relative humidity. This M99 aerosol is one of the most commonly retrieved aerosol types in the global SeaWiFS and MODIS processing. It should be noted that a fixed aerosol model is not the same as the "fixed epsilon" approach often employed in past CZCS reprocessing, as the model accounts for the viewing and solar path geometry dependences of a maritime aerosol.
Aerosol Iteration Scheme
With the aerosol type fixed, and a method developed to estimate water-leaving reflectance at 670 nm (Rrs_670) given water-leaving reflectance at 550 nm (Rrs_550) and chlorophyll (Chl), an iteration process is employed such that (1) a starting Rrs_550 and Chl is assumed, (2) Rrs_670 is estimated and subtracted from the observed reflectance to retrieve an estimate of aerosol reflectance (radiance) at 670, (3) aerosol radiance at 670 is extrapolated to 443, 520, and 550 using the M99 aerosol model, (4) Rrs_550 and chl are recomputed and a new Rrs_670 is estimated. This process is repeated until Rrs_670 stops changing.
Discussion of Approach
While we recognize that the assumption of a fixed aerosol type is not ideal, it must be recognized that CZCS simply lacks sufficient capabilities to independently assess aerosol contributions. Therefore, some assumptions must be made. The scheme described here is as close as possible to the methods currently employed for SeaWiFS and MODIS/Aqua processing within the OBPG, including the iteration scheme, and the aerosol modeling includes full multi-scattering computations using one of the standard models used in the OBPG processing of those modern sensors.
As a verification of this approach, a time-series was generated using SeaWiFS data that was processed using (1) the standard NIR algorithm, and (2) the CZCS algorithm (i.e., treating SeaWiFS as if it lacked NIR bands). Comparisons of normalize water-leaving radiance retrievals at 443nm (nLw_443), retrieved chlorophyll, and retrieved vs modeled nLw_670 are shown below. For this mesotrophic region, the two processing methods yield very similar results.
One possible enhancement to the current OBPG aerosol correction approach would be to utilize a global climatology of aerosol type, such as the SeaWiFS Angstrom fields, to select aerosol type. This could account for "typical" seasonal and open-ocean-to-coastal changes in aerosol type, but it makes the assumption that such large scale variabilities have not changed in 20 years. We may evaluate the performance of such a modification in the future. Another possibility is to use the clearest ocean observations in each orbit, where water-contributions may be predictable at 550 and 670nm, to estimate the spectral change in aerosols at those wavelengths and hence retrieve aerosol type. The aerosol type can then be spatially interpolated to other (not so clear) pixels. This was the approach used in a recent reprocessing by Gregg et al. (2002), wherein epsilon was estimated for the "clear" pixels and interpolated over each scene. Such an approach does assume that epsilon is geometry independent (at least to the extent of the interpolation distance), and it requires that clear pixels exist for each scene. We may evaluate the performance of such a clear-pixel interpolation in the future.
- (1) Carder, K. L., F. R. Chen, Z. P. Lee, S. K. Hawes, D. Kamykowski. (1999)
- "Semianalytic Moderate-Resolution Imaging Spectrometer algorithms for chlorophyll a and absorption with bio-optical domains based on nitrate-depletion temperatures", JGR, Vol 104, No. C3, pp 5403-5422.
- (2) Gould, Jr., R. W., R. A. Arnone, P. M. Martinolich. (1999)
- "Spectral dependence of the scattering coefficient in case 1 and case 2 waters", Appl. Opt., Vol 38, No. 12, pp 2377-2383.
- (3) Gordon, H. R., O. B. Brown, R. H. Evans, J. W. Brown, R. C. Smith, K. S. Baker, and D. K. Clark (1988)
- "A semianalytic radiance model of ocean color", JGR, Vol 93, pp 10,909-10,924
- (4) Bricaud, A., A. Morel, M. Babin, K. Allali, and H. Claustre, (1998)
- "Variations of light absorption by suspended particles with the chlorophyll a concentration in oceanic (case 1) waters: Analysis and implications for bio-optical models", JGR., Vol 103, pp 31,033--31,044
- (5) Gregg, W.W., M.E. Conkright, J.E. O'Reilly, F.S. Patt, M.H. Wang, J.A. Yoder, and N.W. Casey (2002).
- "NOAA-NASA Coastal Zone Color Scanner Reanalysis Effort ," Appl. Opt. 41, 1615-1628.
- (6) Stumpf, R. P., R. A. Arnone, R. W. Gould, Jr., P. M. Martinolich and V. Ransibrahmanakul, (2003)
- "A Partially coupled ocean-atmosphere model for retrieval of water-leaving radiance from SeaWiFS in coastal waters", in "Algorithm Updates for the Fourth SeaWiFS Data Reprocessing", Vol 22, SeaWiFS Postlaunch Technical Report Series.