Premium
A combined radiative transfer model for sea ice, open ocean, and atmosphere
Author(s) -
Fuhrhop Rolf,
Grenfell Thomas C.,
Heygster Georg,
Johnsen KlausPeter,
Schlüssel Peter,
Schrader Meeno,
Simmer Clemens
Publication year - 1998
Publication title -
radio science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 84
eISSN - 1944-799X
pISSN - 0048-6604
DOI - 10.1029/97rs03020
Subject(s) - atmospheric radiative transfer codes , radiative transfer , sea ice , brightness temperature , sea ice thickness , environmental science , sea ice concentration , brightness , climate model , atmosphere (unit) , remote sensing , geology , atmospheric sciences , arctic ice pack , meteorology , climatology , physics , climate change , optics , oceanography
A radiative transfer model to compute brightness temperatures in the microwave frequency range for polar regions including sea ice, open ocean, and atmosphere has been developed and applied to sensitivity studies and retrieval algorithm development. The radiative transfer within sea ice is incorporated according to the “many layer strong fluctuation theory” of Stogryn [1986, 1987] and T. Grenfell [ Winebrenner et al. , 1992]. The reflectivity of the open water is computed with the three‐scale model of Schrader [1995]. Both surface models supply the bistatic scattering coefficients, which define the lower boundary for the atmospheric model. The atmospheric model computes the gaseous absorption by the Liebe et al. [1993] model. Scattering by hydrometeors is determined by Mie or Rayleigh theory. Simulated brightness temperatures have been compared with special sensor microwave imager (SSM/I) observations. The comparison exhibits shortcomings of the ice model for 37 GHz. Applying a simple ad hoc correction at this frequency gives consistent comparison results within the range of observational accuracy. The simulated brightness temperatures show the strong influence of clouds and variations of wind speed over the open ocean, which will affect the sea ice retrieval even for an ice‐covered ocean. Simulated brightness temperatures have been used to train a neural network algorithm for the total sea ice concentration, which accounts for these effects. Sea ice concentrations sensed from the SSM/I data using the network and the NASA sea ice algorithm show systematic differences in dependence on cloudiness.