Jökull - 01.01.2010, Page 55

Upptyppingar seismic swarms
Table 1. Offsets in easting, northing, and depth of ASN locations relative to SIL locations. Relative displacements are
calculated by comparing equivalent events on an event-by-event basis. The catalogue locations represent a comparison of
CMM locations using ASN data with on-line SIL catalogue locations (first column). The large offset in depth is due in
part to different definitions of zero-depth. For CMM, depth is defined as below sea level. For the SIL catalogue locations,
depth is defined as below the nearest one or two seismic stations to an event. The single-event locations are derived by
inverting IMO time picks of SIL data and Cambridge University time picks of ASN data in HypoInverse-2000 (second
column). Offsets are also presented for HypoInverse-2000 location results using Cambridge University and IMO time picks
for six SIL stations common to each network (third column). Overall, the offsets for locations using common stations are
smaller than the offsets produced by network disparities. The offsets are not precisely zero owing to greater location scatter
resulting from the use of fewer stations and also to minor differences in time picking. Lastly, offsets for relative locations
determined by Cambridge University using HypoDD and the IMO using the software developed by Slunga et al. (1995)
are presented (fourth column). These offsets are greater than those obtained by comparing Cambridge University results
to IMO single-event locations relocated in HypoDD. The latter (i.e., +34 m, -344 m, and +649 m in easting, northing, and
depth, respectively) more closely resemble the single-event location offsets (second column). – Mismunur í staðsetningum
með ASN netinu og SIL-kerfinu. Hliðrunin er fundin með því að bera saman reiknaðar staðsetningar hvers skjálfta fyrir
sig. Tölurnar í fyrsta dálknum (Catalogue) sýna muninn á CMM staðsetningum í ASN netinu og daglegum jarðskjálftas-
taðsetningum úr SIL-kerfinu. Munurinn á reiknuðu dýpi er að hluta til vegna mismunandi viðmiðunar. Dýpið sem fengið
er með CMM aðferðinni miðar við sjávarmál, en í SIL-kerfinu er miðað við hæð næstu jarðskjálftastöðva (um 700 metrar).
Í dálki tvö eru bornar saman niðurstöður HypoInverse-2000 staðsetningar, úr SIL-kerfinu og ASN netinu. Í þriðja dálki
eru bornar saman niðurstöður úr HypoInverse-2000 þar sem aðeins eru notaðar 6 SIL stöðvar. Minni heildarmunur er á
staðsetningunum þegar notaðar eru sömu stöðvar og sama staðsetningarforrit. Mismunurinn skýrist af því að munur er á
aflestrunum í kerfunum tveimur og að staðsetningaróvissan er meiri þegar færri stöðvar eru notaðar. Að lokum eru bornar
saman niðurstöður úr afstæðum staðsetningum úr báðum kerfum, þar sem HypoDD er notað á aflestra Cambridgeháskóla
úr ASN netinu, en aðferð Slunga og fleiri (1995) er notuð á SIL gögnin. Meiri munur er á þessum staðsetningum en þeim sem
fást þegar aflestrar úr SIL kerfinu eru notaðir til að endurreikna staðsetningar með HypoDD. Í því tilviki er mismunurinn
+34 metrar, -344 metrar og +649 metrar til austurs, norðurs og í dýpi, sem svipar til samanburðarins í dálki tvö.
Catalogue Single-Event Single-Event (6 SIL) Relative Relocation
Easting (m) -63 +31 -107 -530
Northing (m) -647 -396 +170 -479
Depth (m) +1246 +567 -180 +86
cations, we isolate and invert time-pick data from
six SIL stations (VSH, ADA, HVA, BRU, MKO and
KRE) that are common to both the ASN as well as the
SIL network. The difference here is that the phase on-
set times and pick weights (for identical data from the
same six stations) are determined independently: one
set of time picks is selected by the IMO and the other
set is selected through PPICK at Cambridge Univer-
sity. Each set of picks is passed through HypoInverse-
2000 using the linear-gradient Vatnajökull velocity
model and the single-event location results are pre-
sented in Figure 7.
Albeit scattered, the locations align well in east-
ing, northing, and depth (Table 1), suggesting that
picking precision is not a dominant factor in deter-
mining scatter and offsets in the hypocentres. The
spatial errors for each set of time picks, as estimated
by HypoInverse-2000, are equivalent within 100 me-
tres (i.e., ±0.8 km in the horizontal and ±1.0 km in
the vertical at one standard deviation). The eastward
and southward shifts of the hypocentres relative to lo-
cations derived using the full networks (Figure 6) are
attributed to an asymmetric, one-sided distribution of
the six SIL stations about the earthquakes.
III. Relative Relocations
Finally, we compare SIL relative relocations with
ASN relative relocations in Figure 8. SIL locations
are provided by the IMO using the double-difference
algorithm of Slunga et al. (1995) and ASN locations
are derived using the double-difference algorithm of
JÖKULL No. 60 55