Skip to main content
cancel
 
CERES Home Page CERES Home Page
 
 

Summary of CERES Surface-Only Flux Algorithms (SOFA)


Summary of CERES Surface-Only Flux Algorithms (SOFA).

  • CERES uses several simple surface flux algorithms in addition to the detailed radiative transfer computations in SARB. These are:
  • Why so many algorithms:
    • SW Model A - (Li et al. 1993):
      • Clear-sky only. No retrievals for cloudy footprints. No cloud forcing.
      • Derives surface net SW flux surface from TOA reflected SW flux.
      • Surface insolation computed using surface albedo available only for clear skies.
      • No support currently from the authors. No improvements planned.
    • SW Model B - (Darnell et al. 1992) (Gupta et al. 2001):
      • All-sky model. Derives surface insolation using parameterizations for absorption, scattering, and cloud attenuation.
      • Surface albedo computed internally based on solar zenith angle, surface type, other atmospheric constituents, and cloud condition.
      • Supported by authors in-house at LaRC. Undergoes continuous evaluation and improvements based on product validation.
    • LW Model A - (Inamdar and Ramanathan 1997):
      • No cloudy footprint retrievals. No cloud forcing. Sparse spatial coverage.
      • Based on the premise that OLR and DLR are correlated and the latter can be derived from satellite measurements of the former.
      • Uses CERES window channel (8-12 µm) radiance slated to go away beyond FM-5. Not usable for FM-6 onward.
      • No support currently from the authors. No improvements planned.
    • LW Model B - (Gupta et al. 1992) (Gupta et al. 2010): All-sky.
      • A parameterization based on a detailed radiative transfer computations.
      • Clear-sky flux computed in terms of Ts, lower tropospheric temperature profile, and column water vapor.
      • Cloud effect computed in terms of cloud amount and cloud-base temperature.
      • Supported by authors in-house at LaRC. Undergoes continuous evaluation and improvements based on product validation.
    • LW Model C - (Zhou et al. 2001) (Zhou et al. 2007): All-sky.
      • A parameterization based on a detailed radiative transfer computations.
      • Clear-sky flux computed in terms of Ts and column water vapor.
      • Cloud effect computed in terms of cloud liquid water and ice water paths.
      • Some support available from the primary author.

  • Why so many algorithms:
    • At the beginning of CERES project, there were not many effectively functioning detailed radiative transfer models for surface fluxes while there were several simple algorithms that were in use for some time and already validated.
    • The basic premises underlying these models are very different. Hence, several were introduced one-by-one as checks on each other and also as checks on SARB products.
    • Models A (SW and LW) were introduced first but did not progress beyond clear-sky capability. Therefore, Models B were introduced, both of which are all-sky and had in-house support.
    • These models are extremely fast to run; are well documented in the literature and can be run with a variety of basic inputs. These can be set up by any user and can be used to produce reliable results with minimal resources.
    • LW Model A makes use of the CERES window channel (8-12 µm) radiance that is not expected to be available on future CERES instruments beyond FM-5. LW Model C is being introduced with Edition3 processing to fill the resulting gap and it is an all-sky model.

  • Final thoughts:
    • Models B (SW and LW) are being used in the FLASHFlux project where surface fluxes (in addition to TOA fluxes) are being produced on a near real-time basis (within one week of satellite observations).
    • While it is not possible to state with certainty which of these models are better, Models A (both clear-sky only) are not very useful. Models B are all-sky and will continue to complement the detailed model computed products. LW Model C is a recent introduction and will need to undergo substantial validation.


Model Theoretical Basis Strengths and Weaknesses
SW Model A Surface net SW flux derived first from TOA reflected flux. Surface insolation computed using surface albedo available only for clear skies. No cloudy-sky fluxes; no cloud radiative forcing. No support available from authors. No improvements planned.
SW Model B Surface insolation derived using parameterizations for absorption, scattering, and cloud attenuation. Surface albedo computed internally using solar zenith angle and other physical conditions. All sky model; cloud forcing estimation possible. Strong in-house support available from the authors. Continuously evaluated and improved based on product validation over the entire CERES period. Being used in other projects (GEWEX-SRB and FLASHFlux) at LaRC.
LW Model A Based on the premise of correlation between TOA and surface LW fluxes. Makes use of TOA broadband and CERES window (8-12 µm) measurements. Clear sky only; no cloud radiative forcing. No support available from the authors. No improvements planned. CERES window channel to go away after CERES FM-5. Not usable for FM-6 and beyond.
LW Model B Parameterization based on detailed RTM computations. Derives LW flux using reanalysis meteorology and CERES cloud products. Assumes decoupling between TOA and surface LW fluxes. All sky model; cloud forcing estimation possible. Strong in-house support available from the authors. Continuously evaluated and improved based on product validation over the entire CERES period. Being used in other projects (GEWEX-SRB and FLASHFlux) at LaRC.
LW Model C Parameterization based on detailed RTM computations. Derives clear-sky flux using reanalysis meteorology. Cloud effect derived using cloud LWP and IWP. All sky model; cloud forcing estimation possible. Limited support available from authors. Being introduced with CERES Edition3 processing. Validation still preliminary.

CERES Data Product Information

 
  Image: NASA Logo NASA Official: Dr. Norman Loeb
Page Curator: Edward Kizer
Page Last Modified: 10/16/2017 11:04:36 EST
Site Last Modified: 11/07/2017 17:18:43 EST