the ctbt and advances in earth science
TRANSCRIPT
The Comprehensive Nuclear Test Ban
Treaty (CTBT), opened for signature in
1996, has now been signed by 165
countries and in the words of then UN Secre-
tary General Boutros Boutros-Ghali, it is “a
major milestone on the road to making the
world a safer and more secure place for gener-
ations to come”. The treaty is the culmination
of 40 years of negotiations, characterized by
political and scientific concerns over whether it
could be adequately verified. From the start,
most scientific questions centred on seismology
– vital for verifying a treaty which bans under-
ground nuclear explosions, the predominant
type of nuclear test. This meeting, held in
London on 7–8 February 2002, addressed both
the scientific issues behind the Treaty and the
scientific advances that have resulted.
The first day of the meeting reviewed the pio-
neering research that has allowed a verifiable
CTBT, emphasizing the work of UK seismolo-
gists and particularly those of the Atomic
Weapons Establishment (AWE) at Blacknest,
Berkshire. The Blacknest group was formed
after the initial Geneva 1958 discussions on
nuclear test limitation, to work on “forensic
seismology” – seismology applied to the verifi-
cation of arms control treaties. This small
group has produced significant advances and
published work in this field out of all propor-
tion to its size. Prof. Alan Douglas led the
group from 1982 to his “retirement” in 2001,
and the meeting recognized his considerable
contributions to forensic seismology over
nearly 40 years. Peter Marshall OBE, Alan’s
successor and chief seismological adviser to the
UK side during the treaty negotiations, intro-
duced the meeting on the first day.
R Blandford (AFTAC, USA) gave an Ameri-
can perspective and noted the different, but
complementary approaches US and UK seis-
mologists have taken in solving treaty verifi-
cation problems. He suggested that the most
significant AWE contribution was the recogni-
tion that explosion yield estimates from seis-
mic P waves are influenced by zones of high
and low attenuation near source and receiver,
and related proving studies. These results,
indicating high attenuation of P waves propa-
gating to and from the area of the Nevada
Test Site, were fiercely debated in the 1970s as
they suggested that the US was overestimating
the yield of Soviet tests, but later verified by
jointly conducted experiments at the US and
Soviet test sites.
Seismogram modelling
UK seismologists have long promoted seismo-
gram modelling as an aid to interpretation
and understanding of seismograms in terms of
source processes and wave propagation
through attenuative, inhomogeneous media.
J Hudson (Cambridge) demonstrated how
studies of seismic scattering had developed
from early work on modelling the effects of a
deep river valley near the Eskdalemuir array
to, recently, explanation of the effects of an
explosion cavity on surface-reflected body
waves. The modelling theme was extended by
R Pearce (CTBTO) who showed how a search
for a systematic way of explaining observed
amplitudes of direct P and surface reflected pP
and sP waves led to a powerful method for
seismic source mechanism determination from
only a few well-recorded seismograms. Pearce
showed how negative evidence of pP and sP
may be the most valuable in determining the
source mechanism, and how discrimination
between earthquakes and anomalous artificial
events, such as closely separated explosions, is
possible. Comparison of body and surface
wave magnitudes, mb and Ms, provides a
robust discriminant, but a sparse monitoring
network may generate few Ms measurements,
and these may be biased due to path effects
such as focusing. N Selby (AWE) showed how
station and path corrections for Ms improve
its reliability.
Unusual seismic events provide an incentive
for forensic seismologists to develop and refine
techniques. D Bowers (AWE) examined seismic
signals associated with the sinking of the sub-
marine Kursk and showed evidence of two
underwater explosions of magnitudes 1.5 and
3.5 ML, separated by 135 s. Spectral features
indicative of bubble pulse formation and inter-
ference of direct and surface reflected waves,
indicate that the first explosion (estimated
2–80 kg TNT) was contained at a depth of
60 m or less, whereas the second explosion
(estimated 2–8 tonnes TNT) at a greater depth
(~120 m), opened the hull. Oceanic seismic
events also figured prominently in the review
by B Massinon (LDG, France) of French work
towards verification of the CTBT, which also
included infrasound to detect atmospheric
tests. French studies on the use of seismic
“T”-phases recorded at island seismic stations
have provided an effective, cheaper alternative
to expensive hydrophones for detecting and
locating events at sea. P Marshall (AWE) noted
that the synergy between the seismic, hydro-
acoustic and infrasound technologies in the
global CTBT International Monitoring System
(IMS) of sensors had significantly helped in
making the CTBT possible. C Guralp (Guralp
Systems) provided a fascinating review of how
rapid advances in seismometer development,
including the use of mechanical and electrical
feedback, have allowed the development of
borehole instruments providing data – vital for
analysis of source-signal characteristics and
path attenuation effects.
In his closing presentation, A Douglas (AWE)
noted some unsolved problems, and asked
what had been learned about the Earth from
forensic seismology. In his opinion this was not
as much as expected, considering the potential
for examination of the seismic response close
to large controlled sources, albeit at some cost.
Effects of anelastic attenuation, geometric
spreading, spalling at the surface, reactivation
of existing faults and distribution of down-
ward radiated energy have never been proper-
ly measured, nor an observed time delay in
surface-reflected pP wave satisfactorily
explained. Even now, relatively few test epi-
centre and origin times have been released to
allow the examination of station and network
bias due to heterogeneity in the crust and
upper mantle. He noted how varied research
conducted on a consistent theme invariably
turns out to be useful, and illustrated this with
reference to path-dependent wave attenuation
by describing how observations and modelling
Meeting report
2.33April 2002 Vol 43
The CTBTand advancesin earth scienceDavid Booth and Peter Maguire report on a two-day meeting organized by
the Royal Astronomical Society and the British Geophysical Association,
that showed how scientific challenges generated by the Comprehensive
Nuclear Test Ban Treaty have been addressed, and how they continue to
stimulate advances in earth science.
have explained many apparent anomalies in
observed seismograms, as well as providing
evidence that seismologists possess the capabil-
ity to verify the CTBT by discriminating
between earthquakes and explosions.
The second day of the meeting showed how
the requirement for accurate location of events
for CTBT monitoring requires detailed knowl-
edge of the deep structure of the British Isles,
and how this is being advanced using past and
present research. Accurate location is vital for
verifying the CTBT. A suspicious event cannot
be identified unambiguously as a nuclear
explosion without an on-site inspection, which
is permitted by the CTBT. The search area is
restricted to 1000 km2, and thus the epicentre
must be accurately determined using a dataset
which will be sparse for small events. Source
location is determined from the records of
monitoring networks by determining the point
at which wave travel times best match theoret-
ical travel times derived from the best earth
models available. Treaty signatories, including
the UK, have been asked to provide data which
will allow global structural models to be
refined for local and regional effects at source
and receiver.
Calibrating seismic stations
Calibration of the 170 global IMS seismic
stations is an immense task and several con-
sortia have been formed to meet the challenges
that it creates. K McLaughlin (SAIC/CMR,
USA) outlined the work his consortium is
doing to improve locations in North Africa
and the Mediterranean, Middle East, and
Western Asia. Initial surface-wave models are
being refined using higher-resolution P-wave
studies, and emphasis is placed on statistical
analysis to quantify the improvements made.
Joint Hypocentre Determination, pioneered by
Alan Douglas, is vital for calibration studies, as
is the use of accurate locations from dense
networks and aftershock studies.
Most information on the velocity structure in
the region of the UK has been gained through
controlled-source wide-angle seismic refrac-
tion and reflection profiles by UK and Irish
academic and government institutions over the
past 30 years. The late Brian Jacob and col-
leagues at DIAS, Ireland, have contributed
much to the success of several deep seismic
profiles in and around the British Isles, and
B O’Reilly (DIAS) reviewed this work, show-
ing how successive experiments improved the
resolution of the velocity structure. R England
(Leicester) has compiled results from existing
UK and Irish profiles, most of which are off-
shore, to produce contoured maps of Moho
depth and average crustal P velocity. Interest-
ingly, computer-generated “unbiased” velocity
contours show no obvious correlation with
tectonic terrane boundaries. However, evi-
dence of magmatic underplating is suggested
by high-velocity zones in the lower crust,
which correlate with areas of maximum uplift
in the early Tertiary, later noted also by
N White (Cambridge). Ample evidence for 3-D
heterogeneity requires the capability to model
it, and R Hobbs (Cambridge) demonstrated an
economical, effective “Phase-Screen” method
for modelling wave propagation at narrow
angles through 3-D models, by reducing the
model to a series of slabs. The method has
been used to show the focusing and defocusing
effects of interface topography, and the sensi-
tivity of receiver function analysis to assump-
tions of a local 1-D structure.
A Chadwick (British Geological Survey)
examined Moho two-way travel time (TWTT)
images derived from seismic reflection profiles,
and considered TWTT perturbations in terms
of lateral variations of the velocity–density
relationship in the crust, and of density in the
mantle. It is difficult to differentiate between
these effects without additional data from
shear-wave velocities and 3-D gravity model-
ling, but an upward perturbation of Moho
TWTT under granite appears consistent with a
locally hotter, less dense upper mantle. G Kim-
bell’s (BGS) broad geophysical perspective
examined crustal thickness variation in the NE
Atlantic continental margin, estimated from
topography and an initial model for sedimen-
tary cover, under different assumptions of
isostasy and lithospheric flexure. Allowance
for density variation between continental and
hotter oceanic lithosphere, gives a good match
between theoretical and seismically observed
Moho depth across the continent–ocean
boundary. The morning concluded with
N White’s absorbing study of sedimentary
basin subsidence analysis, demonstrating an
uplift event beneath the British Isles at about
60–55 Ma, centred on the Irish Sea. The results
are consistent with vitrinite reflectance and fis-
sion track data and suggest a major thermal
event occurring beneath this region during the
early Cenozoic.
Three PhD students showed the way ahead in
the afternoon session. L Brana (Bristol) has
mapped the locations of crustal scatterers in
the UK, and near the former Kazakh test site,
using BGS records of teleseismic P waves, and
showed that P to Rayleigh (Rg) wave scattering
predominates. Where P to shear wave scatter-
ing is present, it correlates with topography,
whereas faults generate mainly Rg scattering.
New UK crust and upper-mantle velocity struc-
tures have been generated from teleseismic
receiver functions using three-component UK
seismograms. P Denton, reporting for J Tom-
linson (both Leicester), showed that while
these generally agree with results from nearby
seismic profiles, there is evidence for a local
subcrustal anomalous velocity zone near the
Isle of Man, possibly associated with the
Iapetus Suture. Recent tomographic images of
upper-mantle velocity structure under most of
the UK down to 400 km (see figure 1) were
displayed by S Arrowsmith (Leeds), who has
inverted teleseismic P-wave delay times at UK
seismic stations. There is dramatic correlation
between mantle low velocity anomalies and the
Tertiary dyke swarms and plutonic centres in
the UK, while fast velocity anomalies in North-
ern Scotland are interpreted as signatures of
Caledonian subduction. In east and central
southwestern Britain, high velocities may be
related to possible Caledonian subducted
lithospheric plates, or the presence of a micro-
craton which has remained relatively unde-
formed throughout the Caledonian and
Variscan orogenies.
Finally, I Smith (BGS) noted that stored data
and models should be accessible, understand-
able and protected against the departure or
retirement of staff and storage media. He
described the framework of the BGS Digital
Geoscientific Spatial Model for the UK, which
also addresses such thorny issues as storage of
textual description of model quality and uncer-
tainty, and description of best practice. There is
not so much useful data around, that we can
afford to lose any of it! �
David Booth (British Geological Survey)and PeterMaguire (Leicester Uni.) were joint organizers ofthis meeting. Meeting programme, abstracts, BGSlinks: www.gsrg.nmh.ac.ukCTBT organization: www.ctbto.org AWE Blacknest: www.blacknest.gov.uk
Meeting report
2.34 April 2002 Vol 43
1: 3-D perspective view of volumes from 50 to600 km depth enclosing high (blue) and low (red)P-wave velocity anomalies beneath the BritishIsles, where the surfaces enclose regions of over0.50% higher and lower velocity. The underlyinggrey sphere denotes the 660 km discontinuity.