Geology Reference
In-Depth Information
to appear darker than the surrounding ice in SAR images
due to their relatively smooth surface, even if covered
with thin ice.
The most suitable sensor for the lead detection should
satisfy two criteria. The first is a spatial resolution that
warrants detection of the narrowest leads. The second is
a wide swath that warrants coverage of a wide enough
area. This will facilitate the compilation of lead statistics
at a synoptic scale. A compromise should be struck as
sensors with a wide swath have usually coarse resolution
and vice versa.
An example of lead distribution in the Beaufort Sea
under the action of the Gyre is shown in the AVHRR
images in Figure 9.8. A high‐pressure system over the
western region of the Beaufort Sea in February-March
2013 caused a relatively warm atmospheric temperature
and a series of high southwesterly winds that drove the
Beaufort Sea Gyre. Cracks started to appear in late
January and spread west toward Banks Island. The series
of images shown in the figure demonstrates the effect of
the continuous storm series on exacerbating the initial
fracturing (top image). Ice broke gradually to form large‐
scale bands of leads. By the end of February (the bottom
image), the leads reached the western coast of Banks
Island, approximately 1000 km away from the location of
the original cracks and leads. It is obvious that the major
leads are geometrically shaped, forming a linear or curved
contour that correlated with geostrophic wind as
explained below. It is interesting to note that this area was
ice free in October 2012, which means that most of the ice
shown in Figure 9.8 was almost certainly FY ice. This ice
responds faster than MY ice to wind action by breaking
and drifting. Leads of this magnitude are usually
observed in this region in late spring when ice starts to
break up. The figure provides further evidence to support
the premise of the undergoing changes of the Arctic ice
cap. The ice during that year was probably thinner than
normal.
Barry et al.
[1989] relied on visual inspection of images
to detect large‐scale leads in the Arctic and explore the
correlation of their orientation with wind field. They
used the visible (0.4-1.1
μ
m) and thermal infrared
(8-13
μ
m) imagery data from the Defense Meteorological
Satellite Program (DMSP) to detect leads at 0.6 km reso-
lution based on the contrast between their signature and
the surrounding ice signature. Geostrophic winds were
derived from the daily gridded pressure data provided by
the Arctic Ocean Buoy Program where wind fields are
referenced to a 1° × 5° grid. Qualitatively speaking, the
geometric pattern of the leads was found to be arranged
roughly parallel to the geostrophic wind direction.
Figure 9.9 shows the geometrical pattern of the leads that
are greater than 300 m in width and the corresponding
mean sea level pressure pattern in the Beaufort Sea on
1. 0
0.8
0.6
Avg of SCICEX cruises
AIM:S1:9/03 to 8/04
AIM:S2:9/04 to 9/05
0.4
0.2
ICESat footprint
0.0
0
50
100
150
200
Open water spans (m)
Figure 9.7
Cumulative distribution of the width of leads in the
Arctic from AIM and SCICEX programs (see text). Data were
acquired during summer and winter. The arrow points to the
70 m footprint width of ICESat‐1 [
Kwok et al
., 2009, Fig. 2b,
with permission from AGU].
100 m in width and therefore cannot be detected by
medium‐resolution sensors (with resolution of a few hun-
dred meters). Therefore sensors with finer footprints
(e.g., SAR) are more suitable to detect these leads. The
70 m footprint (across track) of the Ice, Cloud and Land
Elevation satellite (ICESat‐1) that carried a laser altime-
ter (operated from 2003 to 2010) was used for lead detec-
tion in order to infer the freeboard of the ice cover.
ICESat‐2 is scheduled for launch in 2016.
Kwok
[2003]
reported that the widths of less than 20% of Arctic leads
are comparable to the nominal footprint of ICESat‐1.
Leads can be detected in VIS, TIR, or SAR images. In
all cases they appear in the images as long quasi‐linear or
jagged narrow features. Optical systems are suitable for
lead detection because leads and the surrounding ice
areas appear in the image with the same contrast as they
appear in nature (leads are darker due to their low
albedo). In thermal infrared imagery, the contrast is
maintained because of the significantly higher surface
temperature of the lead compared to the surrounding ice
(recall that the TIR emissivity of both materials is almost
equal as explained in section 7.3.3). While VIS channels
can only be used during sunlit periods, they are preferred
over the TIR during summer when the difference between
the ice and water temperatures is reduced. Wind does not
generate as much wave action in leads as it does in OW,
consequently thin ice grows in leads with a smooth sur-
face under relatively steady conditions. This causes leads
