Geology Reference
In-Depth Information
Modeled SSM/I
T
B
for linear combinations of
4 surface types and 12 different atmospheric conditions
Observed SSM/I
T
B
Calculate set of ratios:
R
obs
= {PR
R
(19), PR
R
(85),
Δ
GR}
Calculate pol. and grad. ratios for all concentrations
(0-100) and all 12 atmospheric conditions:
R
mod
= {PR
R
(19), PR
R
(85),
Δ
GR}
Atmosphere index
Find weather-correcte
d ice concentra
tion
by minimizing
δ
R
=
√
(
R
obs
-
R
mod
)
2
?
Ice concentration C
fy
Modeled value R
mod
Ice concentration
C
th
?
Observed value R
obs
Figure 10.12
Schematic diagram illustrating the calculation procedures in the enhanced NASA Team (NT2)
algorithm.
4. For an actual set of
T
b
observation the same three
parameters PR
R
(19), PR
R
(85), and ΔGR are calculated.
These parameters are also expressed in the vector form
R
obs
= [PR
R
(19), PR
R
(85), ΔGR]
obs
.
5. A search is conducted to find the closest
R
mod
value
to the given
R
obs
(bottom right of Figure 10.12). The
criterion used is the minimum error between the observed
and the modeled observations represented as
version of the radiative transfer equation to account
for atmospheric influences on the observations from
the high‐frequency microwave channel (near 90 GHz). It
should be noted that the SEALION algorithm also uses
measurements near 90 GHz to retrieve ice concentration
using a radiative transfer equation that accounts for
atmospheric influences. The ASI algorithm is based on
the familiar notion that the polarization difference is a
function of ice concentration. The measured polarization
difference
P
is related to the ice concentration
C
by the
following equation:
2
RRR
obs
(10.22)
mod
The algorithm makes a decision to generate thin ice or C‐
type ice (in addition to FY ice) based on a threshold value
of ΔGR. If ΔGR > 0.05, modeled data are generated
using emissivity of FY ice and thin ice along with the
atmospheric index. If ΔGR < 0.05, modeled data are
generated using C‐type ice to replace thin ice.
PCP
1
C Pa
(10.23)
si
,
sw c
,
where
P
s
,
i
and
P
s
,
w
are the polarization difference for
ice and water, respectively. The term between brackets
in the RHS represents the surface polarization differ-
ence
P
s
. The term
a
c
represents the atmospheric influ-
ence and is a function of ice concentration.
Svendsen
et al.
[1987] developed the following equation to
determine
a
c
:
C
. The ARTIST Sea Ice (ASI) Algorithm
The Arctic
Radiation and Turbulence Interaction Study (ARTIST)
program was conducted in March and April 1998 in the
area around Svalbard to measure sea ice parameters
using ground‐based and airborne instruments. Based on
the measurements, the ARTIST sea ice (ASI) algorithm
was developed to estimate ice concentration using the
high spatial resolution of the 85 GHz channels of the
SSM/I [
Kaleschke et al.
, 2001]. This algorithm is an
enhanced version of the N90 algorithm developed by
Svendsen et al.
[1987]. It incorporates a simplified
aPe
11
.
e
011
.
(10.24)
c
s
where
τ
is the atmospheric opacity. This equation is valid
for a horizontally stratified atmosphere with an effective
temperature replacing the vertical temperature profile
[
Svendsen et al.
, 1987]. Denoting the atmospherically cor-
rected polarization difference from OW and consolidated
