NEAREST (INTEGRATED OBSERVATIONS FROM NEAR SHORE SOURCES
OF TSUNAMIS: TOWARDS AN EARLY
WARNING SYSTEM)
http://nearest.bo.ismar.cnr.it
NEAREST 2008 Cruise Preliminary Report r/v Urania
1st Aug 2008 - 4th Sept 2008
Carrara G., Matias L., Geissler W.H., D’Oriano F., Lagalante M., Cianchini G.,Chierici F., Cuffaro M.,
Diaconov A., Doormann U., Favali P., Feld C., Gerber H., Innocenzi L., Labahn
E., Langner W., Lo Bue N., Riminucci F., Romsdorf M., Salocchi A.,
Unglert K., Veneruso M., Wolter R.J., Zitellini N.
ISTITUTO DI SCIENZE MARINE -
SEDE DI BOLOGNA
INDEX
1. Introduction: NEAREST project
1.1 Geological setting
1.2 State of the art for the
tsunami detection
2. NEAREST 2008 Cruise
2.1 Objectives
3. First Leg: OBS recovery
3.1 OBS technical description
3.2 Stations recovery
3.3 Data handling
3.4 Preliminary evaluation of
seismic data recorded by the NEAREST OBS’s
3.4.1. Local
seismicity recorded by the land station network
3.4.2
Definition of an initial 1D velocity model for the location of earthquakes in
the NEAREST deployment area
3.4.3 The work
by Gonzalez et al., 1996
3.4.4 The work
by Gutscher et al., 2002
3.4.5 Initial
velocity model proposed
3.4.6 Location
of events using a small set of OBS data
3.4.7 Looking
for local events not reported by the land stations
3.3.8 Strange
signals recorded at OBS21
3.4.9 Delays
recorded on P-wave teleseismic events
3.4.10. Long
term recording of pressure at the sea bottom
4. Second Leg: GEOSTAR and mooring cable recovery
4.1 Technical description of
the Geophysical seafloor observatory and
instruments
4.2 GEOSTAR recovery
operations
4.2.1 Check of MODUS after
recovering
4.2.2
GEOSTAR acoustic check after recovery
4.3 Mooring cable recovery
5. CTD Survey
5.1
Area description
5.2
Sampling and Methods
6. Multibeam
and chirp survey
6.1 Portimao survey map
6.2 Moroccan survey map
6.3 Chirp data example
7. Instruments
7.1 Chirp
7.2 Multibeam
7.3 CTD
8. Daily
report of the cruise
9. References
10. PARTICIPANTS
Cover: Moon eclypse in the
Graphics and maps by : G. Carrara, F. D’Oriano
We want to thank the Master E. Gentile and the
crew of RV Urania for their great and friendly support and professional
handling of the instruments… and for the graffa’s party!
LIST OF FIGURES
Fig. 1 SW Iberian margin pag.
8
Fig. 2 Locations of the recovered OBSs during the
NEAREST-2008 cruise pag.
11
Fig. 3 LOBSTER
pag. 12
Fig. 4 Deck unit 8011M used to communicate with the release
unit of the OBS pag. 14
Fig. 5 LOBSTER recovery pag.
15
Fig. 6 Location of epicentres from the 1st day of
operation of the NEAREST
OBS network till the 31st July 2008, pag.
18
Fig. 7 Location of epicentres closer than
Fig. 8 Location of the profile studied by Gonzalez et al.
(1996) pag.
21
Fig. 9 Location of the IAM-3 MCS profile pag.
22
Fig.10 Final velocity model derived by Gonzalez et al.
(1996) pag.
22
Fig. 11 Location of the profile 16 investigated by Gutscher et
al. (2002) pag. 24
Fig. 12 Location of the SISMAR OBS and the NEAREST network pag. 24
Fig. 13 Final velocity model derived by Gutscher et al.
(2002) pag.
25
Fig. 14 Synthesis of the velocity models pag.
26
Fig. 15 Waveforms and phase picks for 5 OBSs examined pag. 27
Fig.
Fig. 17 Example of 5-station spectrogram displaying one
seismic event pag. 28
Fig. 18 Example of 5-station spectrogram displaying a new
seismic event pag. 29
Fig. 19 Waveforms and phase picks for the 3 OBSs pag.
29
Fig. 20 Epicentre location for the new event identified from
NEAREST OBS pag. 30
Fig. 21 Example of 4-component spectrogram pag.
31
Fig. 22 Waveforms of the local signals identified at OBS21 pag. 31
Fig. 23 Relationship between the mud volcanoes and the
NEAREST OBS pag. 32
Fig. 24 Shot location of the Moundforce cruise,
August/September 2007 pag.
33
Fig. 25 Wide-angle record section for OBS16 recording one
Moundforce profile pag. 33
Fig. 26 Wide-angle record section for OBS18 recording a
Fig. 27 Recordings of a teleseismic event on 5 OBS’s,
vertical component pag. 35
Fig. 28 comparison of P-wave arrivals and theoretical
travel-times pag.
35
Fig. 29 Available MCS profiles in the
arrival times
of teleseismic events pag.
36
Fig. 30 One day recording of pressure data on 5 instruments pag. 37
Fig. 31 GEOSTAR abyssal multiparameter station pag. 39
Fig. 32 The Modus Module pag. 40
Fig. 33 MODUS and the GEOSTAR station before the deployment
2007-08-25 pag. 40
Fig. 34 MODUS approaching a deep-sea station
(CAD-visualization) pag.
41
Fig. 35 MODUS
control units and video monitors in the URANIA lab pag. 42
Fig. 36 MODUS
in the
Fig. 37 GEOSTAR and
buoy site deployment locations pag.
43
Fig. 38 Termination and the bending restrictor before
the Nearest
recovery in August 2008 pag.
44
Fig. 39 GEOSTAR
station – top view pag.
45
Fig. 40 GEOSTAR
station –view from the cone camera 45° pag.
46
Fig. 41 GEOSTAR
station, docking pin –view from the cone camera 45° pag. 46
Fig. 42
MODUS and GEOSTAR station after the recovery 2008-08-17 pag. 47
Fig. 43 View into the inner area of the docking cone
of MODUS pag.
48
Fig. 44
View into the outer area of the docking cone of MODUS pag. 48
Fig. 45
Vertical acceleration, pitch and pull at the winch during the
deployment of GEOSTAR 08/07 pag.
50
Fig. 46
Vertical acceleration, pitch and pull at the winch during the
recovery of GEOSTAR 08/08 pag.
50
Fig. 47 Sketch of acoustic system parameters and settings pag. 51
Fig. 48 Sketch of general status of GEOSTAR pag.
51
Fig. 49 Sketch
of GEOSTAR DACS status pag.
52
Fig. 50: Drift measured through the oscilloscope pag.
53
Fig. 51 - Example of seismic events acquired by GEOSTAR pag.
54
Fig. 52 – Mooring configuration pag.
55
Fig. 53 Mooring recovery pag. 56
Fig. 54 CTD profile
performed during the NEAREST 2007 cruise pag.
57
Fig. 55 Area of the
oceanographic survey around GEOSTAR site. pag.
58
Fig.
Mediterranean water pag.
59
Fig. 57
CTD_Geostar profile of NEAREST 2007 (a)
and NEAREST 2008 (b) pag. 59
Fig. 58 – Multibeam and chirp survey South of Portimao:
navigation tracks pag.
61
Fig. 59 –
Fig. 60 – Multibeam and chirp survey South of Faro: navigation
tracks pag. 62
Fig. 61 – South Faro multibeam data shaded relief pag.
63
Fig. 62 –
Multibeam and chirp survey off
Fig. 63 – Off
Fig. 64– Example of fluid escape along a chirp line located
off moroccan coasts pag. 64
Fig. 65– Example of faulting of the seabed along a chirp line
located off moroccan coasts pag.
65
Fig. 66 – Example of fold and faulting of the seabed along a
chirp line located
off moroccan coasts pag.
65
Fig. 67 – First Leg NEAREST Team pag.
79
Fig. 68 – Second Leg NEAREST Team pag.
79
LIST OF
TABLES
Table 1: List of participant institutions pag.
6
Table 2. Ranging of OBS-16 and OBS21 pag.
13
Table 3. Recovery parameters pag. 15
Table 4. Mean station coordinates pag. 16
Table 5. Recording parameters pag.
17
Table 6: GEOSTAR main sensors pag.
39
Table 7. List of
requests of STATUS and EVENT DATA LOG pag.
52
Table 8. CTD stations pag.
60
Table 9 – Chirp II instrument parameters pag.
66
Table 10 – Multibeam instrument parameters pag.
67
Summary of the
operations
During the NEAREST 2008 Cruise were performed the
following operations:
1) Recovery of 24 OBSs
2) Recovery of the seafloor multiparametric abyssal
station
3) Recovery of the communication buoy’s cable
4) the acquisition of multibeam bathymetric data
offshore
These operations were planned on behalf the European
Project NEAREST designed to identify, characterize and monitor the
large potential tsunami sources located near shore in the
1.
Introduction: NEAREST Project
NEAREST (Integrated observations from Near Shore
Sources of Tsunamis: towards an early warning system) is carried out on behalf
the EU Specific Programme “Integrating and Strengthening the European Research
Area”, Sub-Priority 1.1.6.3, “Global Change and Ecosystems”, Call identifier:
FP6-2005-GLOBAL-4 (OJ
|
1 |
Consiglio
Nazionale delle Ricerche Istituto Scienze
Marine, Dipartimento di Bologna |
|
ISMAR |
|
2 |
Fundação da Faculdade de Ciências da Universidade de Lisboa - Centro de
Geofísica da Universidade de Lisboa |
|
FFCUL |
|
3 |
Consejo Superior de Investigaciones Cientificas –
Unitat de Tecnologia |
|
CSIC |
|
4 |
Alfred-Wegener-Institute fur
Polar-und Meeresforschung |
|
AWI |
|
5 |
Université de Bretagne Occidentale UMR 6358 Domaines Océaniques |
|
UBO-UMR6358 |
|
6 |
Istituto
Nazionale Geofisica e Vulcanologia Roma 2 section –
Marine Unit RIDGE |
|
INGV |
|
7 |
Technische Fachhochschule Berlin - FB VIII - Maschinenbau, Verfahrens- und Umwelttechnik - AG Tiefseesysteme |
|
TFH |
|
8 |
Instituto Andaluz de Geofísica - Universidad De
Granada |
|
UGR |
|
9 |
Instituto de Meteorologia Divisão de Sismologia |
|
IM |
|
10 |
Centre National pour |
|
CNRST |
|
11 |
XISTOS Développement S.A. |
|
XISTOS |
Table 1: List of participant
institutions
NEAREST is a multidisciplinary project devoted
to the study of the tsunami phenomena in its different aspects which can be
summarized as follows: the identification and characterisation of the large
potential tsunami sources located near shore in the Gulf of Cadiz; the
improvement of near real-time detection of earthquake and tsunami signals by a
multiparameter seafloor observatory (Geostar-like station) for the
characterisation of the potential tsunamigenic sources to be used in the
development of an Early Warning System (EWS) Prototype; the improvement of
integrated numerical models enabling more accurate scenarios of tsunami impact
and the production of accurate inundation maps in selected areas of the Algarve
(SW Portugal), highly hit by the 1755 tsunamis.
In this area, highly populated and prone to
devastating earthquakes and tsunamis (e.g., 1755
The methodological approach is
be based on the cross-checking of multiparameter time series, acquired on the
seafloor by 24 broad band Ocean Bottom Seismometers and by a long-term deep-sea
station developed by the INGV partner upgrading the GEOSTAR technology. This
abyssal station is equipped with real-time communication to an onshore main
stations located in
In addition, NEAREST will search for sedimentological
evidence tsunamis records to improve the knowledge on the recurrence time for
extreme events and will try to measure the key parameters for the comprehension
of the tsunami generation mechanisms.
Another aspect investigated by the project is the
improvement of integrated numerical models for the building of more accurate
scenarios of tsunami impact and the production of accurate inundation maps in
selected areas of the
1.1. Geological setting
The SW Iberian Margin is located at the eastern end of
the Azores-Gibraltar-Fracture zone, which, in agreement with the
plate-kinematic reconstructions (Olivet et al. 1996; Srivastava et al.,
1990), is the Eurasia-Africa plate
boundary.
The area could be divided in two main morphotectonic
domains (Tortella et al., 1997): the first between the Gorringe Bank and Cabo
Sao Vicente to the west, and the
Fig. 1 – SW Iberian margin
The first area is characterized by a complex and
irregular topography, dominated by large seamounts, deep abyssal plains, and
massive rises (e.g. Bergeron and Bonnin, 1991; Gràcia et al., 2003a, Terrinha
et al., 2003; Zitellini et al., 2004) such as the Gorringe Bank. The second
area is characterized by a smoother topography and by a prominent NE-SW
trending positive free-air gravity anomaly (Dañobeitia et al., 1999; Gràcia et
al., 2003b).
During the Triassic-Jurassic break-up of Pangea, the
eastward drifting of Africa respect to Iberia led to the formation of a rift basins between the new continental
margins; this divergent stage ended in early Late Cretaceous. Subsequent
northwards migration of Africa with respect to
From Tertiary up to now the main compression direction has rotated
anticlockwise, currently the latest GPS kinematic models (Nocquet et al.,
2004), show a WNW-ESE main direction of the relative movements between the
African and Iberian plates.
Plate convergence of 4 mm/yr (Argus et al., 1989;
Nocquet et al., 2004) is accommodated, in this area, over a wide and diffuse
deformation zone (Sartori et al., 1994; Hayward et al., 1999) characterized by
significant and widespread seismic activity (e.g., Grimison and Chen, 1986).
This tectonically active deformation zone was been source of the largest
earthquakes that affected the
1.2. State of the Art for Tsunami Detection
The reason for developing
a real-time, deep ocean tsunami measurement system was to foreseen the impact
of tsunamis on coastal areas in time to save lives and protect property.
The first approach to
Tsunami waves monitoring was a combination of tide gauges and seismometers.
After that, in order to provide a much earlier warning of an approaching
tsunami, NOAA (National Oceanic and Atmospheric Administration), developed the
research project for Deep-ocean Assessment and Reporting of Tsunami (DART),
using buoys in deep sea, acoustically linked to sea-floor pressure gauges. In
turn, the buoys would relay the sensor data to a central land site by satellite
radio links.
The first-generation DART was based on an
automatic detection and reporting algorithm triggered by a threshold
wave-height value. The DART II design incorporated two-way communications that
enables tsunami data transmission on demand, independent of the automatic
algorithm.
Each DART gage was designed to detect and report
tsunamis on its own, without instructions from land. The tsunami detection
algorithm developed in the gage's software works by firstly estimating the
amplitudes of the pressure fluctuations within the tsunami frequency band and
then testing these amplitudes against a threshold value. The amplitudes are
computed by subtracting predicted pressures from the observations, in which the
predictions closely match the tides and lower frequency fluctuations. The
predictions are updated every 15 seconds, which is the sampling interval of the
DART gages. The detection threshold was defined using statistical analysis on
background oceanic noise. Based on past observations, a reasonable threshold for
the North Pacific was fixed to
2. NEAREST 2008
CRUISE
2.1 Objectives
The scientific survey was performed
between 1th August and 4th September 2007 offshore Cabo Sao Vicente and in the
3. First Leg: OBS recovery
A passive seismic
experiment was conducted in the
Figure 2. Locations of the recovered OBS during NEAREST-2008
cruise.
Tectonic
structures, in the transition from the Azores fracture zone to the postulated
subduction zone in the area of the Strait of Gibraltar, that have the potential
to cause Tsunamis will be localized and characterized. For this purpose 24
broadband ocean bottom seismometers (OBS) from the German DEPAS instrument pool
coordinated by the Alfred Wegener Institute for Polar and Marine Research, Bremerhaven and the GeoForschungsZentrum, Potsdam
3.1 OBS technical description
24 DEPAS LOBSTER (Longterm
Ocean Bottom Seismometer for Tsunami and Earthquake
Research, see figure 3) K/MT 510 manufactured by K.U.M. Umwelt- und
Meerestechnik Kiel GmbH, Germany, were used during the experiment. They were
equipped with a Güralp CMG-40T broadband seismometer incorporated in titanium
pressure housing, a hydrophone, and a GEOLON MCS (Marine Compact Seismocorder)
data logger from SEND
Figure
3 - LOBSTER
3.2 Station recovery
Approaching to the deployment positions we tried
to range the OBSs at the seafloor to get a better control on their positions
(coordinates, depth). For shallow OBSs (2000 to
Table 2. Ranging of OBS-16 and
OBS21.
Unfortunately, we were not able to locate the
remaining OBS, not only because of the tied time table which became even more
tied due to unplanned stop at Faro due to the need of desembaking one member of
the team because injured and bad sea conditions. We had big problems to hear
the answers from the release units even for station shallower than
Unknown regular signals (one ping about every 10
seconds) were recorded as answers at the deck unit at stations OBS23 and later
at OBS14 (Later on, disappeared at station OBS14 during the ranging measurements).
Figure 4. Deck unit 8011M used to
communicate with the release unit of the OBS.
After we were sure about the arrival of the OBS
at surface the ship was positioned some hundreds meters away from the expected
position of recovery against the waves (normally SSE of the stations). At the
estimated time the OBS could be quickly located at the surface. In most cases
this was done by eye, since the flag and the float units could be easily seen
during the day. Only one flag was missing. The flash lights helped during the
night; also the radio beacons worked well but were not needed to locate the
OBS.
The recovery on deck was possible in most cases
within 10 to15 min after the appearance of the OBS at the sea surface due to
the good work of the bridge and the deck crew. During the recovery an entering
hook and the backboard crane were used. The position of recovery on deck was
taken to calculate the mean coordinate of the OBS at depth from deployment and
recovery coordinates. In most cases the difference in coordinates between
deployment and recovery is very small (table 3 and 4)
.
Figure
5. LOBSTER recovery.
Table 3. Recovery parameters.
Table 4. Mean station coordinates.
Immediately after recovery of the OBS on deck we
tried to manually stop the recording and synchronize the internal clock with
GPS time signal using SENDCOM-3 interface. With only few exceptions, all stations stopped recording already before
recovery due to the following reasons: “disk full” and “battery low”. The first
cause was expected due to the length of the period of operation and the high
sampling rate. All recorders of the 9 stations with full disks stopped
recording properly and allowed a GPS synchronisation without any problems. The
battery low at 9 stations was unexpected, since the capacity of the 132 Li
cells was estimated to be enough for 12 month recording. The battery charge was
so low (with only one exception), that
even there was not enough power left to keep the clocks running (although there
is a safety mode, which worked for one station). Therefore the recorder lost the
synchronisation and we were not able to get the time shift of the internal
clock.
After stopping the recording, the stations were
cleaned and dismounted. All removable components were stored in boxes, the
LOBSTER itself were stacked and stored onboard. With the exception of two
stations (OBS16, OBS21) the stations were in very good conditions after
recovery. These two stations showed corrosion at the power connectors of the
recorder tube. Similar corrosion was maybe a cause of the failure of one power connector
during the first deployment cruise in 2007. There was almost no cover with
carbonates as it was observed before in the
3.3 Data
handling
After cleaning, the recorder tube was taken off
from the tube and connected to the Desktop PC by FireWire. Data retrieval from
MCS recorders was performed using send2x software (mcscopy). Afterwards the raw data were decompressed into s2x format
using mcsread. During this stage the
time correction (if possible) was done. The final stage of first data
conversion on the ship was the conversion into mseed format (seedwrite) to allow the very first
quality check of the data. Two copies of the raw data were saved on external
hard disks (less than 500 GB disk space), a third one on DL tapes. Furthermore,
one copy of the uncompressed s2x-files (altogether more than 1 TB) and two
copies of mseed data (500 GB) were save on external hard disks.
During checking some data we observed that maybe
not all seismometers (at least components) levelled well. This will be checked
together with other parameters (quality control) during the next weeks at AWI
Institute. In all cases the hydrophone seems to work properly. We expect that
there is no severe loss of data and therefore will not certainly hamper the
analysis and interpretation of the data. We also hope that the timing problem
of 9 stations (missing synchronization) could be solved with the use of the
seismic signals itself.
Table 5. Recording
parameters.
We want to thank the captain and the crew of RV Urania
for their good and friendly support and professional handling of the
instruments.
3.4 Preliminary
evaluation of seismic data recorded by the NEAREST OBS’s
3.4.1. Local
seismicity recorded by the land station network
The first evaluation of the seismic data
recorded by the NEAREST OBS network is provided by the comparison with the
local and regional seismicity recorded by the land station network. Fernando
Carrilho from IM (Instituto de Meteorologia) provided a list of all events
recorded by his institute from the first day of recording, the 29th August 2007
till the 31st July 2008, the beginning of the cruise. In the study area, from
33.5ºN to 38.5ºN and from 14ºW to 5ºW, this dataset comprises 1893 seismic
events that include also quarry blasts on land and some not well-located
events. The epicentres are shown in fig. 6, together with the location of the
NEAREST OBS network and known seismic stations operating in the area.
Figure 6 – Location of epicentres from the 1st day of
operation of the NEAREST OBS network till the 31st July 2008, as provided by
Fernando Carrilho (IM). Also shown are the known recording seismic stations
operating on land
The recordings provided by the land stations
will be very useful for the full exploitation of the NEAREST dataset, so that a
list of the known seismic stations and their major characteristics is provided
as an EXCEL file as Annex CD-SP1.
From the large dataset shown in fig. 6, we
selected as first priority for the investigation of the NEAREST OBS network area
all events that were at least at
Fig. 7 – Location of 276 selected epicentres, which
are closer than
To assess the data quality recorded we screened
all 257 events that should have been recorded by OBS18, considering the
selected events set in fig7. The mini-seed data was converted to Nordic waveform
data and the waveforms were examined using the SEISAN Seismic Analysis package
(Havskov and Ottemöller, 2005). We classified the events as Very Good, when we
have clear P and S readings and the polarity of the P phase can be seen without
ambiguity, Good, when we have clear P and S phases, Fair, when we have clear S
but P is doubtful, Unknown, when a small signal is seen but the phase cannot be
clearly identified, and Null when no signal could be seen. The results of the
screening are shown in the table below.
|
Quality |
Rating |
||
|
Very Good |
17% |
52% |
72% |
|
Good |
35% |
||
|
Fair |
20% |
|
|
|
Unknown |
7% |
27% |
|
|
Null |
20% |
|
|
We have 52% of events that are fully usable for
location, 17% of which will also contribute for the determination of source
mechanism using P-wave polarity. Only 27% are expected to be useless for
location. We must say that these statistics may not represent a typical value.
A preliminary screening on OBS16, on top of a thick sedimentary layer, showed
less optimistic values. Considering that different instruments will have
recorded different events with good quality, it may be expected that at least
2/3 of the 276 events recorded in the area by the land seismic network will be
well relocated by the NEAREST OBS network.
3.4.2 Definition of
an initial 1D velocity model for the location of earthquakes in the NEAREST
deployment area
Using the data recorded and recovered onboard we
proceed to try to locate some of the seismic events identified by the land
network. In order to do this we must define an a priori velocity model for the
study area.
The location of the events by the land network,
computed at the Instituto de Meteorologia, uses a general velocity model that
is accepted for all
|
Vp (km/s) |
Depth (km) |
|
6.10 |
0.0 |
|
6.40 |
11.0 |
|
6.90 |
24.0 |
|
8.20 |
31.0 |
|
8.50 |
90.0 |
|
Vp/Vs=1.75 |
|
This velocity model seems not appropriate for the
location of local earthquakes since it doesn’t show a sedimentary layer and the
crust is
3.4.3 The work by
Gonzalez et al., 1996
This work presents a velocity model along a profile
that extends from
Fig. 8 – Location of the profile
studied by Gonzalez et al. (1996), reproduced from their fig1.
The velocity model in Gonzalez et al. (1996) was
obtained integrating information from the IAM-3 MCS line, wide-angle recordings
on 10 land stations and gravity modelling. It does not provide a direct measure
of the seismic velocities in the area but it should be considered as a good
indication. Their final model is shown in fig.10.
Fig. 9 – Location of the IAM-3 MCS profile used in the work by
Gonzalez et al. (1996) and the NEAREST network.
Fig.10 – Final velocity model derived by Gonzalez et al. (1996).
The layer characteristics described by Gonzalez at al.
(1996) are presented in the table below:
|
Layer |
Description |
Vp (km/s) |
|
1 |
Water |
1.5 |
|
2 |
Upper sediments |
2.2 |
|
3 |
Lower sediments |
3.7 |
|
4 |
Upper crust |
5.8 to 6.0 |
|
5 |
Middle crust |
6.4 |
|
6 |
Lower crust |
6.8 to 6.9 |
|
7 |
Upper mantle |
7.8 to 7.9 |
We will consider as representative to our study area
two velocity profiles extracted from Gonzalez et al. (1996) model, one at
The velocity models are presented in the table below:
|
@ |
@ |
||
|
Vp (km/s) |
Depth (km) |
Vp (km/s) |
Depth (km) |
|
1.5 |
0.00 |
1.5 |
0.00 |
|
2.2 |
3.18 |
2.2 |
4.41 |
|
3.7 |
4.69 |
3.7 |
6.77 |
|
5.8 – 6.0 |
5.62 |
5.8 to 6.0 |
9.80 |
|
6.4 |
13.06 |
6.4 |
12.63 |
|
7.8 – 7.9 |
15.43 |
6.8 to 6.9 |
14.79 |
3.4.4 The work by Gutscher
et al., 2002
In this paper Gutscher et al. present the results of
refraction and wide-angle reflection modelling from coincident MCS and OBS
recordings. The location of the profile studied is shown in fig. 11 (from the
original paper). The relation between this profile and the NEAREST network is
shown in fig.12.
Fig. 11 – Location of the profile 16
investigated by Gutscher et al. (2002).
Here reproduced from their fig1.
We will consider as representative to our study area
the 1D velocity profile that is inferred below the westernmost OBS, at
Fig. 12 – Location of the SISMAR-16
MCS profile and OBS’s used in the work by Gutscher et al. (2002) and the
NEAREST network.
Fig. 13 – Final velocity model
derived by Gutscher et al. (2002) reproduced
from the original fig3.
The velocity profile inferred from the colour plot in Gutscher et al.
(2002) work is shown in the table below:
|
Vp (km/s) |
Depth (km) |
Layer |
|
1.5 |
0.00 |
Water |
|
1.8 – 3.0 |
2.74 |
Upper sediments |
|
3.7 – 4.2 |
6.87 |
Lower sediments |
|
5.2 – 6.8 |
10.34 |
Crust |
|
7.9 – 8.1 |
17.09 |
Upper mantle |
3.4.5 Initial
velocity model proposed
The models presented above are all shown in fig.14 as
velocity-depth profiles. We may see that the models proposed by Gutscher et al.
(2002) and the Gonzalez et al. (1996) @
|
Vp (km/s) |
Depth (km) |
Layer |
|
2.2 |
0.0 |
Upper sediments |
|
3.8 |
6.5 |
Lower sediments |
|
5.8 |
10.0 |
Upper crust |
|
6.5 |
13.0 |
Lower crust |
|
7.9 |
16.0 |
Upper mantle |
|
8.1 |
20.0 |
|
|
8.3 |
50.0 |
|
Fig. 14 – Synthesis of the velocity models presented: G2002 – Gutscher
et al. (2002); IM – Instituto de Meteorologia;
G1996-90 – Gonzalez et al. (1996) @90 km; G1996-140 – Gonzalez et al. (1996) @140 km. The proposed model is shown in pink.
3.4.6 Location of
events using a small set of OBS data
Using the velocity model described in the
previous paragraph and the deployment coordinates of the NEAREST OBS network,
we used SEISAN to locate a few of the events that belonged to the list provided
by IM. At the time of this exercise we had only data from 5 instruments, OBS16,
OBS18, OBS19, OBS21 and OBS22 (this with uncertain absolute time). The
waveforms and phase picks are shown in fig.15 for the event at 21/2/2008 20:39
with magnitude ML=2.6
Fig. 15 – Waveforms and phase picks for 5 OBS’s examined, event
20080221-20:39 ML=2.6.
The epicentre location of this event, using only OBS data, is shown in
Fig11 as a red circle. When comparing this location to the original one
provided by IM we see that, despite the poor distribution of the stations, the
OBS location is pretty good.
Fig. 16 – In red, epicentre obtained
by the analysis of 5 OBS’s, event 20080221-20:39 ML=2.6. In black we see the
original location provided by IM.
3.4.7 Looking for
local events not reported by the land stations
One of the added values of the NEAREST OBS
network is the ability to detect and locate local events that cannot be
identified and analysed by the land seismic network. The fully exploitation of
the acquired dataset will imply a very long examination of the complete records
for each station. To facilitate this task we propose the use of spectrogram
images that summarize in a smaller single file all the information from the
network over one day.
We present next the spectrogram analysis applied
to the datasets comprising the recordings of 5 instruments, OBS16, OBS18,
OBS19, OBS21 and OBS22. In fig. 17 we show a piece of spectrogram showing an
earthquake that is well recorded on all 5 instruments. The horizontal scale is
time, one tick per minute. The vertical scale is frequency, from 1 to 25 Hz.
The colours display the spectral energy on each frequency bin red being the
greatest. One earthquake is characterized in the spectrogram by a vertical
alignment of energy. To facilitate the use of this utility, we include in the
spectrogram one vertical green line for the time of one event that is already
known. In this example the seismic land network already located this
earthquake. When we have a vertical alignment without the green line then we
have a completely new event. This is illustrated in fig. 18.
We see also in fig13 that not all the OBS’s
recorded the event and the amplitude varied from instrument to instrument. For
the event identified in fig.18 we
were able to pick P and S phases on 3 instruments and make a location. The
waveforms and picks are shown in fig.19,
while the location is shown in fig.20.
Fig. 17 – Example of 5-station
spectrogram (vertical components) displaying one seismic event that was already
in the land network event list.
Fig. 18 – Example of 5-station
spectrogram (vertical components) displaying one completely new seismic event.
Fig. 19 – Waveforms and phase picks
for the 3 OBS’s examined for the new event identified 20080115-01:15.
Fig. 20 – Epicentre location for the
new event identified by 3 sensors from
the NEAREST OBS network.
The new event is located in the Horseshoe
Abyssal plain, in a area surrounded by other OBS’s, meaning that a better
location will be obtained using the complete dataset.
We performed the examination of 5-station
spectrograms for 5 consecutive days. On average, for this dataset, we found a
new event per day. However only one could be well located by the 5 instruments,
all others lacked a sufficient number of phases. This preliminary evaluation
gives a great expectation in relation to the full dataset. A significant number
of new events, more than 100, are expected to have been recorded by the NEAREST
OBS network.
3.3.8 Strange
signals recorded at OBS21
We can use the spectrogram analysis also to
single station signals, to investigate signals that are very local and are not
recorded in other instruments. While doing the testing of this methodology we
found that OBS21 reported a frequent signals without network significance. An
example of spectrogram is shown in fig.21.
We may not that, contrary to earthquakes, the signals on the seismometer have
no correspondence in the hydrophone. This may indicate a very local source
where the energy is not enough to make a good coupling of the surface movement
to the water column. The waveforms that correspond to the spectrogram in fig16
are shown in fig.22.
Fig. 21 – Example of 4-component
spectrogram for OBS21 displaying a series of unique signals (from bottom to
top: H, Z, Y, X components).
Fig. 22 – Waveforms of the local
signals identified at OBS21.
Since the OBS21 is very close to one mud volcano
(fig.23) Wolfram Geissler suggested that the activity recorded by the seismic
sensor maybe related to the circulation of fluids in the area. This is one
topic that should be investigated further while screening the complete dataset.
Fig. 23 – Relationship between the
mud volcanoes in the
and the NEAREST OBS network.
3.4.9 Wide angle recordings of the Moundforce
cruise, August/September 2007
During the NEAREST deployment cruise in 2007
there was another cruise in the
Fig. 24 – Location of the shots for
MCS acquisition during the Moundforce cruise, August/September 2007.
Fig. 25 – Wide-angle record section
for OBS16 recording the closest
Fig. 26 – Wide-angle record section
for OBS18 recording the
We may see that the OBS16, locate on top of the
accretionary wedge, over a thick pile of sediments, did not record any crustal
phase while OBS18, located in a more favourable geological environment,
recorded some week crustal arrivals and a week PMP phase reflected in the
crust-mantle boundary.
It is recommended that the seg-y files from the
recorded Moundforce profiles are prepared and distributed to the interested
partners as part of the data delivery package from the NEAREST OBS survey.
Furthermore it is recommended that the Moundforce research group be contacted
to foster the future investigation of joint MCS and wide-angle profiles.
3.4.9 Delays
recorded on P-wave teleseismic events
The LOBSTER sensors recorded teleseismic events
that can be used, for example, on P-wave tomography to investigate the mantle
structure. One example of a record of a teleseismic event, Z-component, on
OBS16, OBS18, OBS19, OBS21 and OBS22 is shown in fig.27.
If we compare the observed travel-times with
theoretical ones computed from an average earth velocity model, we see that
each stations shows a delay. This delay is systematic on several records and
maybe related to the deep structure of the earth. But to be sure of this, a
correction must be applied to the arrival times that takes into account the
thickness of the sediments and crust below the seismic station. Fortunately a
number of MCS profiles are available in the
Fig. 27 – Recordings of a
teleseismic event on 5 OBS’s, vertical component.
Fig. 28 – comparison of P-wave
arrivals and theoretical travel-times.
Fig. 29 – Available MCS profiles in
the
correcting arrival times of teleseismic
events.
3.4.10. Long term
recording of pressure at the sea bottom
Wolfram Geissler reported some questions regarding
the recording of the hydrophone data, suspecting that the very broad range of
the sensor was clipped at the low frequencies by the data acquisition system.
Also, some very low frequency fluctuations were observed on some of the data
files. To illustrate these questions we show in fig.30 one daylong recording of pressure data for 5 instruments.
Fig. 30 – One day recording of
pressure data on 5 instruments.
We may see that 3 of the sensors recorded one
low-frequency sinusoidal fluctuation that could be due to tidal effects,
showing that these instruments recorded all the frequency spectrum of the
signal. Two other sensors record some faster fluctuations that seem to
propagate from one instrument to the other. These fluctuations maybe due to
atmospheric pressure variations. This could be confirmed by correlating the
records with atmospheric pressure measured on land stations. Variations between
instruments will have to be checked with all the dataset available.
4. SECOND
LEG: Geostar and mooring cable recovery
The main target of this leg was the recovery of
the abyssal station and the mooring cable. The buoy was already recovered on 20
October 2007 after a failure of the mooring cable (see bouy operation report
Oct. 17-21; Nov 23-27, 2007 available on the NEAREST web site http://nearest.bo.ismar.cnr.it/
). Additional targets of the second leg were the completion of the swath
bathymetric mapping of the continental slope along the Maroccan margin, between
4.1
Technical description of the Geophysical
seafloor observatory and instruments
The Geostar system is a single-frame autonomous
seafloor observatory able to collect multiparameter data with a unique time
reference for long-term investigations.
The technology of this observatory derives from
the synergy among research institutes and industries starting from 1995 to
develop seafloor systems able to operate from shallow water up to deep sea.
During these years, a fleet of observatories has
been built on behalf the funding support of European Commission (i.e. GEOSTAR;
GEOSTAR-2; SN-1;ORION-GEOSTAR-3; ASSEM etc.), bringing more and more
improvements at the main technology of benthic observatories. These systems satisfy the main conditions of
seafloor observatories: multidisciplinary, long-term monitoring, unique time
reference, autonomy, and development of (near) real-time communication system
for warning of local events.
The last generation of Geostar seafloor
observatory, planned in the framework of NEAREST project, is equipped with:
a) geophysical and oceanographic sensor
package
b) central acquisition, control unit (central
clock)
c) data processing unit
d) local memory storage
e) acoustic communication system
All of these characteristics are required to be
able to acquire scientific multiparametric data, to detect real-time events
(seismic and water pressure) and to communicate possible warning messages.
The observatory is constituted by three main
sub-systems:
a) Bottom station constituting
the monitoring system (fig.31)
b) MODUS vehicle that allows
deployment and recovery procedures (fig.32)
c) Buoy system representing the
communication system
Fig. 31 - GEOSTAR abyssal
multiparameter station
The Bottom station consists in a marine aluminum
frame hosting instrumental sensor packages (see table 6), compass controlling
heading, pitch and roll of the observatory during the deployment, lithium
batteries for power supply, echosounder to determine the distance between
Geostar and bottom surface during the deployment, electronic for data
acquisition, hard disks for data storage, underwater part of acoustic
communication system.
|
Sensor |
Sampling
rate |
Acquisition |
|
3-comp. broad band Seismometer |
100 Hz |
Continuous |
|
3-comp. Accelerometer |
100 Hz |
Continuous |
|
Hydrophone |
100 Hz |
Continuous |
|
Pressure sensor |
1 smp/15 sec |
Continuous |
|
Gravity meter |
1Hz |
Continuous |
|
CTD & Transmissometer |
1 smp/min |
Continuous |
|
ADCP |
1 smp/ hour |
Continuous |
|
3 comp. Currentmeter |
5 Hz |
Continuous |
Table 6: GEOSTAR main sensors
The acquisition data is entirely controlled by a
central unit (DACS: Data Acquisition and Control System) that prepares and
updates the hourly data messages, performs the TDA algorithm and transmit data messages on request; also it is
able to send in real-time warning messages of detected events towards the
surface communication system (buoy).
DACS manages a wide set of data having quite
different sampling rate (from 100 Hz to 1 sample/15 sec), tagging each datum
according to a unique time reference set by a central high-precision clock.
Fig. 32 – The Modus Module
MODUS (MObile Docker for Underwater Sciences) is used
within NEAREST project to recover the GEOSTAR station that has been deployed
during the first campaign of the project in late August 2007, Fig. 33.
Fig. 33
-MODUS and the GEOSTAR station before the deployment 2007-08-25
The general concept is shown in Fig. 34. MODUS is
connected to an umbilical providing the needed power and the glass fibres for
the exchange of data for controls, video signals, sonar signals and other
sensors and tools. Moreover, the umbilical carries the entire load of MODUS
(8000 N in water) plus its payload, such us the GEOSTAR station (15000 N in
water) and the force induced by its own weight (17900 N / km). A controlled
SONAR head, with the ability to detect objects of the size of the station in a
distance of 150 –
The set up of MODUS took place on the URANIA while
being in the
Fig. 34
-MODUS approaching a deep-sea station (CAD-visualization)
Fig. 35 -
MODUS control units and video monitors in the URANIA lab
Fig. 36 -
MODUS in the
4.2 GEOSTAR
Recovery operations
The Urania vessel reached the operational area
on August 15, at about 18:00 (fig.37).
The sea state was not considered calm enough to start the operation for the
recovery of the abyssal station. Based on the weather forecast, it was decided
to wait for the following day before to start the operations. During the night
five CDT stations were performed around the location of the abyssal station
(see chapter 6).
Fig. 37 - GEOSTAR and buoy site deployment locations.
The official GEOSTAR and Buoy anchorage coordinates are respectively : 36°21.887’N 9°28.874’W and 36°22.056 N - 009°28.88 W.
On August 16, at 12:00 the sea conditions were
considered good enough to start the recovery of the abyssal station. The mean
wind was 15 knots from NW and the mean wave high was
No special occurrence until a water depth of
about
Return to the surface 2008-08-16
17:32
System recovers step by step 2008-08-16 18:02 ERRORs occur only
rarely
System fully recovered 2008-08-16
18:27 ERROR
free, fully operational
System on board 2008-08-16 19:04 END
of dive
The system check afterwards gave no indication
of damage on the fibre lines. ODTR measurements have been conducted for this.
The results showed the same situation as before the dive. In consequence there
is no increase of dB-loss because of damage due to operational procedures
during this dive. Nevertheless, a signal attenuation occurred, this likely to
the increase of tension in the cable and eventual to the bending at the
termination with the loose bending restrictor Fig.38, which only can be assumed
but not confirmed.
Fig. 38 -
Termination and the bending restrictor before the Nearest recovery in August
2008
Before diving again all FO-connectors were
cleaned and reconnected.
At 19.44 (UTC) the vessel arrived on the GEOSTAR
deployment site and the first operation planned was the acoustic interrogation
of the seafloor observatory through ATS-V acoustic modem. First of all, the acoustic link with the
underwater modem (ATS-V-USS) was checked in order to request the GEOSTAR
station status. All the communication attempts with the underwater acoustic
failed.
The day after (2008-08-17 ) we tryed again the
recovery starting at 09:00 local
time and we record the first errors at a depth of
Complete loss of Sonar 2008-08-17 09:59
Altimeter random operation
2008-08-17 10:09
Distance to the sea floor
Video remains, rest is “lost”
2008-08-17 10:15
Search for GEOSTAR begins with the positioning of the vessel only,
unable to move or turn MODUS, as the thrusters are non operational because of
the failure of data transmission.
GEOSTAR on video
2008-08-17 10:43
GEOSTAR lost 2008-08-17 10:55
Master is tuning in for the search
GEOSTAR on video
2008-08-17 11:20
Master improves positioning significantly, so we remain close to the
station.
Station so close
2008-08-17 11:36
Station closer
2008-08-17 11:39
Docking
2008-08-17 11:40
Return to the surface
2008-08-17 11:41
System recovers step by step 2008-08-17
12:22
System fully recovered 2008-08-17 12:44
fully operational
System on board 2008-08-17 13:23 END
of dive Fig. 42
Fig. 39 -
GEOSTAR station – top view from the vertical orientated stern camera
Fig. 40-
GEOSTAR station –view from the cone camera 45°
Fig. 41 GEOSTAR station, docking pin –view from the
cone camera 45°
Fig.
42 MODUS and GEOSTAR station after the recovery 2008-08-17
During the deployment we had a
smart collision that can be verified on the video documentation, which caused a
nice bump on the cone, Figs. 43 and 45.
Fig.
43 - View into the inner area of the docking cone of MODUS with the traces of
docking
Fig. 44-
View into the outer area of the docking cone of MODUS with the traces of
docking
The repetition of the errors of the data transmission
during the deployment can not be explained explicitly. The facts are the
following:
- MODUS system is fully operational before the deployments
- MODUS system is fully operational after the deployments
- Transmission errors at the depth of
- Video uplink is operational, at full depth
- Recovery of the data transmission without any error indication at
- Obviously there is a hysteresis in the entire system, as the
reoccurrence of the signal takes place at a lower depth as the loss. The
cause of the difference seems to disappear while MODUS is on deck, or it
is a real hysteresis loop for unknown reason.
4.2.1 Check of
MODUS after recovering
Processing the log files from the deployment
cruise in August 2007 provides the following chart, Fig. 45. No indication
of data losses or any other special occurrences. Pull (decreases after
deployment of the station- DEP), pitch comes to a more or less constant value
(usually there is a small gap between the station and MODUS, so slight changes
are likely), acceleration is constant. After releasing the station [REL] and
lifting the data start to change with faster changes induced by the surface
waves à vessel à umbilical à MODUS. These data are consistent.
During the recovery dive different effects could
be monitored as shown in the graph, Fig. 46. Pull is constantly running as
these signals come from the sheave of the winch directly to the onboard
computer of MODUS. Pitch and acceleration are interrupted frequently. The
deeper it goes the more data are lost. Significant and surprising is that in a
period of lower pull (less amplitude overlay) the signals more or less
disappear completely and reappear with higher pull amplitudes. Nevertheless,
the transmission is interrupted in the end completely as already described
above. The entire data set and the cable will be investigated carefully after
our return.
REL DEP
Fig. 45 - Vertical acceleration, pitch and pull at the
winch during the deployment of GEOSTAR 08/07
Fig. 46 - Vertical acceleration, pitch and pull at the
winch during the recovery of GEOSTAR 08/08
4.2.2
GEOSTAR acoustic check after recovery
On
08/18/2008 after recovery, GEOSTAR was washed and a second series of
“Head-to-Head” acoustic tests (consisting in direct contact between the surface
and underwater transducers) were performed on the deck in order to check the
acoustic functionalities and get STATUS and DATA messages from GEOSTAR DACS
(Data Acquisition and Control System). By means of the HMI-NEAREST program, the UPLOAD ALL command was given in order to
get settings and parameters of the underwater acoustic modem (fig. 47). In this
case, the link was successful and the system replied correctly.
Fig.47: sketch of acoustic system
parameters and settings showing a normal status of ATS-V-USS
Then,
the DACS STATUS command was sent to GEOSTAR through HMI-NEAREST program in order to obtain information of the general
status of the station as shown in the fig.48:
Fig. 48: Sketch of
general status of GEOSTAR
The most relevant
anomaly was a delay of about 34h with respect to the current one.
In order to
understand the reasons of this delay, a complete check was performed (see next
paragraph).
On 08/19/2008 a
Direct Serial Link and Log File Download was carried out in order to
gather more information, Log File were downloaded by a direct cable link to
GEOSTAR. A request of DACS status was sent to confirm the correct
functionality of the system that suddenly replied (fig.49).
Fig. 49 – sketch
of GEOSTAR DACS status
A decrease
in the Voltage (2 Volts less in comparison with previous DACS STATUS request)
was noted. This fact could suggest a not-normal general status of the battery
package, probably caused by the low temperature (~2 °C) in the Atlantic
seafloor. The following table 7 shows a list of requests of STATUS and EVENT
DATA LOG file directly downloaded via serial link port in order to better
investigate the functionality of GEOSTAR
during the deployment period.
|
Requested Date |
Result |
Requested Date |
Result |
|
16
Oct 2007 |
Files
Found |
16
Mar 2008 |
Files
Found |
|
16
Nov 2007 |
“ |
16
Apr 2008 |
“ |
|
16
Dec 2007 |
“ |
16
May 2008 |
“ |
|
16
Jan 2008 |
“ |
16
Jun 2008 |
“ |
|
16
Feb 2008 |
“ |
16
Jun 2008 |
“ |
|
|
Check Date |
Result |
|
|
|
1
Aug 2008 |
Files
Not Found |
|
Table 7. List of requests of STATUS and EVENT DATA LOG
The
system shutdown was scheduled for July 27th 2008 and the result of
these interrogations shows that before the 27th of July the files
are present while on August 1st 2008 are not as expected.
Successively
it was decided to disassembly the Orca
Rubidium clock in order to quantify its status and drift.
At
16.55 (UTC), the power cable was removed from the DACS vessel and a voltage of
28.35 V was measured. After that, the DACS was taken off the vessel. The drift
measurement was done before the programmed clock inter-calibration, in order to
avoid the complete discharge of two 9V safety batteries caused by high power
request during this operation. A direct link via serial port allowed a
date-time request: the answer was August, 19th 20:15:21, the current
time. Then, the drift was measured linking the clock to the GPS antenna, and a
drift of 184 ms over almost 1 year was found. The clock behaved as expected
based on the experience made in previous experiments. The following figure
(fig.50) shows the display of the oscilloscope used to measure.
Fig. 50: Drift
measured through the oscilloscope
At the
end of the drift measurement, the inter-calibration of the clock ended and the
clock was powered by an external supply in order to assure a stable voltage.
After
the drift measurement, the three Flash Cards in the DACS were temporarily
removed to backup the data:
- MCU (Master Control Unit): 4 MB;
- DAU (Data Acquisition
Unit): 2.35 GB;
- SDU (Seismometer Data Unit): 257 MB.
Although
the environmental low temperature produced a decreasing of the battery
efficiency, the system was able to acquire correctly until July 5th. We have to underline that, for safety
reasons, a shutdown of the system was programmed by July 27th. A
further analysis on the hard disk showed a 44 GB of memory load.
Fig 51
- Example of seismic events acquired by GEOSTAR.
4.3 Mooring
cable recovery
The mooring scheme deployed in the previous NEAREST2007 cruise
is shown in the following figure:
Fig 52 – Mooring configuration
The mooring of the buoy failed on October 19th
2007 due to the breaking of the stainless steel cable located just beyond the
floating body, and the buoy started drifting in open ocean. The mooring line
was lost on the seafloor including the acoustic transducer which is expected to
lie on the seafloor in the nearby of the observatory.
The recovery operations were performed in 21 of
August 2008 when the weather conditions permitted the work in the area (fig.
53). Approaching the position of 36°22.056
N - 009°28.88 W, on a water depth of
3225 mt a first tentative to communicate with the acoustic
releaser was carried out without success. Only after one hour the system
started to answer, then the release signal was sent. After 70 minutes the releaser
was at 170 mt below sea level. At that point the Master of Urania decided to
grab the cable, using a
n. 1 acoustic release,
n. 6 buoys Nautilus
about 1000 mt of polypropilene cable
n. 1 swivel
the remaining components of the mooring were lost.
Fig. 53: Mooring recovery
5. CTD Survey
In order to characterise, from oceanographyc
point of view, the deployment site of GEOSTAR an hydrological survey is
necessary to define a microscale dynamic of the water masses involved in the
5.1 Area
description
The Gulf of Cadiz area plays a crucial role
being involved in the Mediterranean Water (MW) outflow that, flowing parallel
to the topography, reaches its density equilibrium after the 7°W and it
separates in two cores located at different depth: UMW (Upper Mediterranean
Water) centred at 700-
