Jökull - 01.01.2009, Blaðsíða 96

Sigurjón Jónsson
sponds well with the area that was found to be moving
from GPS measurements and therefore shows that the
Þófi deposits became active well before the first GPS
measurements were carried out.
Profiles across the moving area show better that
the moving area is only 200 m broad and has LOS
displacement of up to 2 cm (Figure 3), while up-
slope profiles indicate that the moving area is 100–
150 m above sea level. The local topographic slope
of Þófi at this location is around 14◦ yielding a local
incidence angle of 37◦ (Figure 3). This implies that
2/ sin(37o) = 3.3 cm of surface parallel down-slope
motion is needed to explain the observed 2 cm LOS
displacement, or an average velocity of 9 cm/year dur-
ing the 4.5 months the interferograms spans.
Interferograms from 1999 show a similar pattern.
The best interferogram spans six months from May
to November 1999 and although it is noisy, it clearly
shows signs of consistent phase change in Þófi (Fig-
ure 4d). In this case the deformation appears to be
slightly stronger than in the 1998 interferogram and
the area that is moving is larger. The center of the
moving area appears decorrelated, possibly due to too
much displacement, but the amount of LOS range in-
crease during these six months exceeds 2–3 cm (7–
10 cm/year creep velocity assuming surface parallel
motion).
No data from this track exist from the year 2000
and in early 2001 the ERS-2 satellite started malfunc-
tioning. Therefore, no suitable ERS data of Þófi ex-
ist that coincide with the GPS observations made in
2001–2004. However, two Envisat interferograms are
shown spanning 70 days in 2004 and 35 days in 2005
(Figure 4e-f) and neither of them exhibit signs of dis-
placements in Þófi. The same is true for other inter-
ferograms from 2004 and 2005. This result agrees
with GPS observations that show no systematic dis-
placements of measurement points during 2004–2005
(Tómas Jóhannesson, pers. comm., 2007).
VOPNAFJÖRÐUR
A broad valley extends inland from the Vopna-
fjörður fjord to the southwest with the Krossavíkur-
fjöll and Smjörfjöll mountains on the southeastern
side that are over 1200 m high while the hills to the
northwest are less than 200 m high (Figure 1). The
flat valley-floor is a home to several farms, particu-
larly on the southeastern side of the valley.
An interferogram of Vopnafjörður spanning 35
days during the summer of 1995 shows clearly move-
ments on the northwestern slopes of Krossavíkurfjöll
(Figure 5a and 6b), between the rivers of Skjald-
þingsstaðaá and Haugsá, above the farm of Svína-
bakkar. These deposits were not known to be active
before it was here detected by InSAR, although they
had been described in the literature (Guðmundsson,
1995b). The moving deposits are at elevation of 2–
500 m and appear to be more than 1 km broad near
its top (where the image becomes decorrelated due
to snow), but somewhat narrower further down-slope.
The southern margin is sharp as indicated by the dis-
continuities in the interferogram. The displacement
amounts to at least 6 cm of LOS range increase im-
plying about 10 cm downward displacement (about
1 m/year) of the deposits, assuming slope-parallel mo-
tion. The pattern of displacement is, however, rather
complicated and includes some internal discontinu-
ities, which shows that the deposits do not move as
a coherent mass.
Due to interferometric decorrelation and the rela-
tively high deformation rate of these deposits, inter-
ferograms that have short time spans provide more in-
formation. Figure 6c-d shows two interferograms that
span 3.5 months and 5 months in 1997 and 1998, re-
spectively. Both show clear signs of slope movement,
but the observed displacements are small. The 5-
month interferogram is somewhat decorrelated, prob-
ably due to snow, but the 3.5 month interferogram
exhibits clearly a maximum LOS displacement of
only about 2 cm (12 cm/year assuming surface par-
allel velocity) and that the lower part of the deposits
does not move at all. An Envisat interferogram that
spans 35 days in 2005 shows no displacements (Fig-
ures 6e), while a one-year 2004–2005 interferogram
clearly indicates several cm of displacements (Fig-
ure 6f). Comparison with the 35-day interferogram
from 1995 (Figure 6b) therefore demonstrates that the
slope motion at this site is far from being steady in
time.
96 JÖKULL No. 59