Jökull - 01.01.2009, Page 100

Sigurjón Jónsson
tively creeping show that this technique can be useful
in monitoring unstable slopes at many locations and
in detecting sites where slope creep is active.
In this study I concentrated on ERS-1 and ERS-
2 data from 1993–1999, which all are from a sin-
gle frame on descending track 424, and on ascending
and descending Envisat data from 2004–2005. More
ERS data from track 424 exist that could be used for
a more complete InSAR analysis for the time period
1992–1999 and some of the data from after 2001 may
also prove to be usable. Data from descending tracks
195 and 152 may also provide additional information.
Furthermore, data of this area from other satellites
exist, mainly from Radarsat-1, JERS-1, and ALOS,
but the archives of data from these satellites were not
investigated. The Canadian Radarsat-1 satellite was
launched in 1995 and is still in orbit. It operates on
a similar wavelength as the European ERS and En-
visat satellites and does probably not provide much
extra information about the slope deformation during
time periods when good ERS and Envisat data exist.
However, there may be useful data from 2000–2004
in the Radarsat-1 archives and there may exist more
ascending data than in the ERS/Envisat archives. The
Japanese JERS-1 satellite was in orbit during 1992–
1998 and operated on L-band wavelength of 23.5 cm,
which is much longer than the C-band wavelength
(5.6 cm) used by the other satellites. The advantage
of this longer wavelength is that temporal decorrela-
tion due to surface changes and snow is less promi-
nent, so longer-term interferograms using data from
this satellite, i.e. spanning several years, would likely
prove to be useful. However, JERS-1 radar data of
Iceland have never been actively used for interferom-
etry, despite undoubtedly rich data archives.
The results from Seyðisfjörður and Vopnafjörður
show that movement of unstable slopes in eastern Ice-
land is episodic, with periods where they do not move
at all, while during other periods they may move by
many cm. This episodic behavior demonstrates the
importance of continuous monitoring of Icelandic
slopes, as slow-moving sites may suddenly accelerate
or dormant deposits may be reactivated. Such mon-
itoring is possible to certain extent with the current
radar satellites such as Envisat, i.e. in a similar way
as was done within this project using 2004–2005 En-
visat data. However, it is important to realize that data
above many regions are not routinely acquired.
CONCLUSIONS
The coherence of C-band radar interferometry is
generally good enough for single- and multi-month
observations of slope movement in East Iceland.
Inter-annual and multi-year observations can also pro-
vide useful information on larger deposits, but not
on the smaller sites as some filtering is generally re-
quired, which prevents small-scale analysis. The main
InSAR measurement problems in East Iceland are
first, the lack of a digital elevation model of adequate
quality, which prevents analysis of long-baseline in-
terferograms, and second, the frequent high-elevation
snow cover, which often limits the use of spring and
autumn images.
Deformation was detected in Þófi near Seyðis-
fjörður in InSAR data from 2 years before 2000, when
surface cracks were discovered in the field. Interfer-
ograms spanning several months in summers of 1998
and 1999 show 2–3 cm of LOS displacement that may
represent surface parallel velocities of 7–10 cm/year.
This rate is lower than the maximum observed GPS
velocities during 2001–2002 (up to 33 cm/year), but
no suitable InSAR data exist from that time period
to compare with the GPS data. No deformation was
detected in InSAR images 1995–1997 or 2004–2005
at the Þófi site, indicating that the deposits creep
episodically.
More than 10 locations of previously unknown
slope creep were discovered in East Iceland in the pro-
cessed interferograms. In some cases the locations of
observed creep correspond to deposits that have been
geomorphologically documented. The most promi-
nent moving deposits were found in Vopnafjörður and
in Loðmundarfjörður. The Vopnafjörður creep shows
variations in both displacement rate and areal extent
during the observation period, which further demon-
strates the episodic behavior of creeping slopes in East
Iceland.
100 JÖKULL No. 59