Dg. Barber et Ef. Ledrew, ON THE LINKS BETWEEN MICROWAVE AND SOLAR WAVELENGTH INTERACTIONS WITHSNOW-COVERED 1ST-YEAR SEA-ICE, Arctic, 47(3), 1994, pp. 298-309
Electromagnetic (EM) energy at solar and microwavelengths will interac
t with a snow-covered sea ice volume as a function of its geophysical
properties. The seasonal metamorphosis of the snow cover modulates the
relative distribution of the three main interaction mechanisms of EM
energy: reflection, transmission, and absorption. We use a combination
of modeling and observational data to illustrate how the total relati
ve scattering cross section (sigma(o)) at microwavelengths can be used
to estimate the surface climatological shortwave albedo and the trans
mitted Photosynthetically Active Radiation (PAR) for a snow-covered, f
irst-year sea ice volume typical of the Canadian Arctic. Modeling resu
lts indicate that both 5.3 and 9.25 GHz frequencies, at HH polarizatio
n and incidence angles of 20-degrees, 30-degrees, and 40-degrees can b
e used to estimate the daily averaged integrated climatological albedo
(alpha). The models at 5.3 GHz, HH polarization, at 20-degrees, 30-de
grees, and 40-degrees incidence angles were equally precise in predict
ions of alpha. The models at 9.25 GHz were slightly less precise, part
icularly at the 40-degrees incidence angle. The reduction in precision
at the 40-degrees incidence angle was attributed to the increased sen
sitivity at both 5.3 and 9.25 GHz to the snow surface scattering term
(sigma-degree(ss)) used in computation of the total relative scatterin
g cross section (sigma-degree). Prediction of subsnow PAR was also pos
sible using the same combination of microwave sensor variables utilize
d in prediction of alpha, but because subice algal communities have ev
olved to be low light sensitive, the majority of the growth cycle occu
rs prior to significant changes in sigma-degree. A method of remote es
timation of snow thickness is required to be scientifically useful. Ob
servational data from the European ERS-1 SAR were used to confirm the
appropriateness of the modeled relationships between sigma-degree, alp
ha, and PAR. Over a time series spanning all conditions used in the mo
deled relationships, the same general patterns were observed between s
igma-degree, alpha, and PAR.