Stic"and 
a proc- 
chemis- 
Infor- 
LS V RS 
  
  
  
MICROWAVE REMOTE SENSING OF SEA ICE 
R.K. HAWKINS, A.L. GRAY, and C.E. LIVINGSTONE, L.D ARSENAULT 
Canada Centre for Remote Sensing (CCRS) 
2464 Sheffield Road, Ottawa, Ontario, Canada, KlA OY7 
ABSTRACT 
This paper reviews a series of airborne microwave experiments on sea 
ice conducted by the Canada Centre for Remote Sensing during the period 1979 to 
1982. Principal sensors used include a Ku-band, dual-polarized scatterometer, a 
K-band radiometer and the SAR-580, a synthetic aperture radar operating at X-, 
C-, and L-bands with dual polarization. The potential of simultaneous active 
and passive sensors for the detection and identification of various sea-ice 
classes is examined with profiling sensor data and the implication for 
operational reconnaissance is supported by radar imagery examples. The 
influence of radar parameters, including frequency, polarization, incidence 
angle, resolution, and sensitivity, on the interpretability of microwave imagery 
of sea ice is discussed for airborne and satellite systems. 
Introduction 
The need to meet developmental and navigational requirements involving 
detailed large area surveillance of the northern coastal regions of Canada where 
sea-ice hazards are significant is well documented (Lapp & Lapp, 1981). It is 
also clear that many of the requirements must be met with remote sensing 
techniques which can provide all-weather, day or night coverage. Microwave 
frequencies are the most plausible candidates because of their relative ease of 
progagation through the atmosphere. This paper explores the use of both active 
and passive microwave sensors for the detection and identification of sea ice, 
first reviewing the seasonal variation of the active and passive microwave 
signatures of sea ice and then discussing the role of the fundamental parameters 
in defining the utility of an imaging radar for sea-ice work. 
Some background into sea-ice characteristics is required to understand 
the direction and intention of this research. Generally, because of its 
inherent salinity, sea ice varies not only in thickness but in mechanical and 
physicial properties as well. The WMO scheme (Dunbar, 1969) for identifying and 
characterizing sea ice according to its age, thickness and form is well 
recognized and provides a convenient basis for ice descriptions. Table I is a 
brief summary of the scheme. 
Using passive techniques; Wilheit et al., in 1972 demonstrated that 
within a range of frequencies old ice could be separated from younger forms 
using emissivity from airborne experiments. Gloerson et al. (1973) also used 
brightness temperature measurements to confirm these results and Meeks 
et al. (1975) were able to study five ice classes with different polarizations 
and look angles. Troy et al. (1981) reviewed several studies before presenting 
their work from the Greenland Sea. They concluded that FY, MY, and open water 
can be reliably distinguished with passive techniques at a single frequency and, 
with additional frequencies, it is possible to even determine ice concentrations 
for situations when sensor footprints contain several ice classes and water, as 
outlined in earlier work by Tooma et al. (1975). Satellite-borne imaging 
radiometers like the ESMR's on Nimbus V and VI and the SMMR's on SEASAT and 
Nimbus VII have been used to provide ice edge position, ice concentration and, 
under cold conditions, ice type fractions for both Arctic and Antarctic sea-ice 
cover (e.g. Gloersen et al. 1982) 
789 
he is Sa RE | -—