Jökull - 01.01.2010, Page 56

Martens et al.
7209
7210
7211
7212
7213
7214
N
o
rt
h
in
g
[k
m
]
437
438
439
440
441
442
E
a
s
ti
n
g
[k
m
]
20Jul 12:00 21Jul 12:00 22Jul 12:00 23Jul 12:00 24Jul 12:00
13
14
15
16
17
18
D
e
p
th
[k
m
]
Date in 2007
Figure 6. ASN single-event locations (red) with respect to SIL-system single-event locations (blue) in easting, northing,
and depth. Here, travel times from each network are inverted using HypoInverse-2000 and the Vatnajökull velocity model
(Figure 4). Furthermore, all locations have been re-datumed to depth below sea level. The time picks for the ASN are
selected at Cambridge University using PPICK (Figure 3). The time picks for the SIL-system are selected at the IMO using
similar manual refinement procedures. As with the catalogue locations (Figure 5), systematic offsets in northing and depth
persist (Table 1). – ASN staðsetningar á stökum skjálftum með komutímaaflestrum úr PPICK forritinu (mynd 3) (rauðir
punktar), ásamt staðsetningum á stökum skjálftum með handvirkt leiðréttum komutímum úr SIL-kerfinu (bláir punktar). Í
báðum tilfellum er notað staðsetningarforritið HypoInverse-2000 og Vatnajökuls-hraðalíkanið (4. mynd). Dýpið hefur verið
leiðrétt, miðað er við sjávarmál. Eins og á 5. mynd má sjá kerfisbundna hliðrun til norðurs og í dýpi (tafla 1).
Waldhauser and Ellsworth (2000) (i.e., HypoDD). All
locations are redatumed to depth below sea level.
Each dataset exhibits well-defined, tightly clus-
tered locations. It is interesting, however, that the
offset in depth from the single-event locations (Fig-
ure 6) has all but disappeared (Table 1, second and
fourth columns). Furthermore, the offset in easting
has shifted by over 0.5 km with respect to the single-
event locations. Here, the SIL relative locations have
been provided directly by the IMO. When the SIL
system time picks used to derive the single-event lo-
cations, however, are subsequently run through Hy-
poDD, we do not find such a dramatic shift in the off-
sets. Indeed, we derive offsets that are very similar
to those obtained when computing single-event loca-
tions.
IV. 3D Spatial Distributions
We now examine the distribution of hypocentres in
three dimensions (3D) for SIL catalogue locations,
SIL relative locations, and ASN relative locations by
fitting the full dataset of 288 hypocentres to a plane
using linear regression. Figures 9, 10 and 11 present
the results looking along the strike direction of the
56 JÖKULL No. 60