Geology Reference
In-Depth Information
ice. The model results are shown in Figure 8.28. The
reduction in albedo can be significant at high particulate
concentrations. Both clean ice and light pond show peak
values at 475 nm. All albedo values were calculated using
direct illumination at a zenith angle of 50°. The results do
not vary significantly from values calculated for diffuse
incident irradiance.
roughness, salinity, temperature, porosity, and brine
pocket geometries and distributions. In the presence of
snow, the emissivity becomes affected by the snow depth,
density, wetness, grain size, and any frozen layers within
the snow pack (e.g., crust or ice lenses). Other composites
on ice surface such as slush, freezing rain, and superim-
posed ice will further complicate the determination of
the emissivity.
Microwave emissivity determines the emitted radiation
from the emitting layer (penetration depth of the signal).
The depth of this layer in sea ice varies between a few mil-
limeters and decimeters, depending on its physical compo-
sition and the microwave frequency (section 8.5). The
temperature of the layer is difficult to determine and in
most studies has been assumed to be equal to the surface
temperature (determined from TIR data). But this assump-
tion could lead to errors in estimating the emissivity.
Many studies have been conducted to estimate emissiv-
ity of sea ice in the microwave frequencies.
Comiso
[1983]
conducted a global‐scale investigation of sea ice emis-
sivity using both infrared and microwave observations
from SMMR.
Hewison and English
[1999] retrieved sur-
face emissivity of snow and ice over the Baltic Sea in the
frequency range 24-157 GHz.
Haggerty and Curry
[2001]
examined the variability of sea ice emissivity estimated
from airborne passive microwave measurements at 37,
89, 150, and 220 GHz near the Surface Heat Budget
of the Arctic Ocean (SHEBA) ice camp.
Mathew et al
.
[2008] retrieved emissivity of FY and MY ice in the
Arctic using observations from the Advanced Microwave
Sounding Unit on the polar‐orbiting satellites NOAA‐15,
NOAA‐16, and NOAA‐17.
Three approaches are commonly used to estimate the
emissivity of snow‐covered sea ice. The first approach
uses in situ measurements of brightness temperature from
well-characterized ice surface (in the presence or absence
of snow). Measurements are usually conducted using
surface‐based radiometers [
Grenfell and Comiso,
1986;
Comiso et al.
, 1989;
Grenfell
, 1992;
Shokr et al.
, 2009].
In the second approach the brightness temperatures are
obtained from airborne or satellite‐borne measurements
[
Grenfell and Lohanick,
1985;
Haggerty and Curry
, 2001;
Mätzler
, 2005;
Mathew et al
., 2008, 2009;
Harlow,
2011].
In these two approaches calculation of emissivity proceeds
based on a simple radiative transfer equation. In the third
approach the emissivity is calculated from a model that
employs ice thickness and a few physical parameters [
Mills
and Heygster,
2011]. A brief account of each approach is
presented, followed by results from selected studies. The
focus is placed on passive microwave bands that are com-
monly used for the retrieval of sea ice parameters (e.g., ice
concentration, thickness, and snow depth).
Surface‐based radiometers (SBR) are commonly used
to measure brightness temperature of the emitted
8.4. eMissivity data in the MicRowave Band
In the case of a composite material such as sea ice,
emissivity becomes an indirect function of the physical
temperature because the latter affects the composition.
For example, brine volume and brine pocket shape and
spacing vary with ice temperature, and those factors
affect the composite emissivity of sea ice (brine salinity
affects the absorption and brine pocket shape affects scat-
tering of the microwave signal). section 4.5.2 includes
information on geometrical characteristics of brine pock-
ets. Another example is about the snow whose composi-
tion is affected by temperature and therefore its emissivity
becomes a function of the temperature‐driven composi-
tion. Snow wetness, grain size, and ice layering within the
snow pack are all driven by temperature. This section pre-
sents an outline of commonly used methods to determine
emissivity from passive microwave data, followed by sam-
ples of data on emissivity from sea ice and snow. Only the
emissivity in the microwave range is presented because
the emissivity of ice, snow, and OW are practically stable
in the TIR band regardless of the composition as men-
tioned before.
A rough estimate of sea ice surface emissivity can be
inferred from the relations between emissivity and Fresnel
reflection [equations (7.18) and (7.19)]. When an incident
wave falls normal to the surface with refractive index
n
t
,
the Fresnel reflection becomes equal for both polariza-
tions and can be obtained from equations (7.4) or (7.5)
after substituting
θ
n
=
θ
t
= 0 and
n
i
= 1:
2
1
1
n
n
(8.18)
The Fresnel equation is applicable to the young ice and
FY ice because the observed reflection or emission origi-
nates mainly from a thin saline layer at the snow‐ice inter-
face [
Zwally et al.
, 1983]. Using a refractive index of 1.8
in equation (8.18), which is typical for sea ice according
to
Vant et al.
[1974], the resulting emissivity is 0.9184.
This approximate value falls within the wide range of
microwave emissivity of snow‐covered sea ice. An accu-
rate estimate of emissivity, however, requires accounting
for a few factors pertaining to the surface and the emit-
ting layer of the ice and snow. These include surface
