Jökull - 01.01.2016, Page 42

Jonathan L. Carrivick et al.
Figure 9. Analysis of surface horizontal motion by manual feature-tracking of a common line of features, such
as the end of major crevasses. Usage of these ‘lines’ mitigated problems with finding exact points to track.
Annotations denote most major horizontal shifts only for clarity and have units of metres since previous image.
Note that only changes within a year could be resolved: between years horizontal displacement was too large to
confidently see the same feature. The location of these panels is indicated in Figure 1. – Mæling á íshreyfingu
með handvirkri greiningu á hliðrun línulegra yfirborðsauðkenna s.s. sprunguenda. Línur auðkenna auðvelda
mælingu á hliðrun þegar erfitt er að finna ákveðin staðbundin auðkenni til þess að fylgja. Merkingar sýna
lárétta hliðrun í metrum miðað við fyrri mynd á nokkrum stöðum. Athugið að hliðrun milli mynda var einungis
unnt að ákvarða innan sama árs. Hliðrun milli ára var of mikil til þess að unnt væri að koma auga á sömu
auðkenni með vissu. Staðsetning myndanna er sýnd á 1. mynd.
For comparison, application of a simple ‘flux gate’
calculation, whereby the ice volume discharged from
the reservoir to the receiving areas is compared with
the product the velocity of ice, ν, flowing through
a cross-section with an area, A (for a parabola A =
2/3×width×depth) over the duration of the surge, T
(here assumed to be one year), in the vicinity of the net
zero elevation change isoline (Figure 2C), produces a
depth-averaged and cross-sectionally averaged veloc-
ity, ν, of ∼200 m yr−1. This is only an indication of
the order-of-magnitude of the velocity. The ice-flow
velocities could have been much higher for shorter
periods.
Runoff in the Jökulsá á Fjöllum river, which in-
cludes input from Kverkfjöll, was statistically higher
(ANoVA p value <0.01) in the years 2010 to 2013
compared to 2008, 2009 and 2014 (Figure 10). In
contrast, the Kreppa discharge, though visually raised
was not statistically higher. The difference between
the runoff recorded at the two sites in those two sets
of years is even more obvious if the volume of water is
considered by integrating the discharge through time.
42 JÖKULL No. 66, 2016