1 International Data Centre Preparatory Commission for the Comprehensive Nuclear-Test- Ban Treaty Organization Processing of the new Processing of the new Infrasound Arrays on the IDC Infrasound Arrays on the IDC Testbed - Initial Experience Testbed - Initial Experience Nicolas Brachet Acoustic officer, CTBTO/IDC/SM Preparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization Provisional Technical Secretariat Vienna International Centre P.O.Box 1200 A-1400 Vienna AUSTRIA [email protected]Presented at the Infrasound workshop in Kona, Hawaii November 2001
22
Embed
International Data Centre Preparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization 1 Processing of the new Infrasound Arrays on.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Processing of the new Infrasound Arrays Processing of the new Infrasound Arrays on the IDC Testbed - Initial Experienceon the IDC Testbed - Initial Experience
Nicolas BrachetAcoustic officer, CTBTO/IDC/SM
Preparatory Commission for theComprehensive Nuclear-Test-Ban Treaty OrganizationProvisional Technical SecretariatVienna International CentreP.O.Box 1200A-1400 [email protected]
Presented at the Infrasound workshop in Kona, HawaiiNovember 2001
2 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Outline of the talkOutline of the talk
• Data received and processed at the IDC
• Infrasound processing at the IDC
• Goals and motivations for improvement of the system
• Case studies
• Problems in the processing of infrasound data at the IDC
3 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Incorporation of new IMS stations in IDC processingIncorporation of new IMS stations in IDC processing
• IDC is using two processing pipelines in parallel• Operations: automatic and interactive analysis performed on a daily
basis - generates the IDC products for distribution• Testbed: mirror of Operations used as a testing environment (software
development, software configuration, integration of new stations in the processing)
• In 2001, IDC is processing the data from 6 IMS infra stations• IS07 (Australia): September 2001 (Testbed), October 2001 (Operations)• IS08 (Bolivia): August 2001 (Testbed), November 2001 (approved for
ops)• IS10 (Canada): September 2001 (Testbed), October 2001 (Operations)• IS26 (Germany): April 2001 (Testbed), October 2001 (Operations)• IS34 (Mongolia): August 2001 (Testbed)• IS57 (USA): October 2001 (Testbed)
4 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Goals and motivationsGoals and motivations
• Improvement of the know-how in infrasound processing• Need for a better understanding of the signals
Establishment of a catalog of reference events
A few case studies presented in this talk
• Becoming familiar with the latest methods in infrasound processingAssessment of existing tools used at the IDC
New developments in infrasound signal processing
(in 2001, cooperation with French NDC: PMCC is being tested at the IDC)
• Station tuning• Currently at the IDC: standard configuration used for all stations• Need to include station specifics in the processing
Assessment of the detection capability for each station
5 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Reference events collected in 2001 at the IDCReference events collected in 2001 at the IDC
•Pacific bolide (23 Apr 2001)
•Earthquakes in Peru (23 Jun, 26 Jun, 7 Jul 2001)
•Eruption of the Etna volcano, Italy (Jul 2001)
•Explosion at a chemical factory in Toulouse, France (20 Sep 2001)
•Mining activity in Canada (11 Sep 2001)
6 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Case study: Pacific bolide (23-Apr-2001)Case study: Pacific bolide (23-Apr-2001)
• PMCC produced the same type of results as for June 23 event
• Libinfra-PIDC detected only the seismic arrivals
Comparison of results for REB event in Peru (07 Jul 2001)
Note: phases identified as Noise by StaPro (cannot be associated to an automatic event)
11 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Case study: seismic event in Peru - Case study: seismic event in Peru - 07 Jul 2001 (Cont.)07 Jul 2001 (Cont.)
1
2
3
3
2
Polar diagram of the infra arrivals
seismic infra1 21’ 3
11’
1’Ms=7.5
IS08
REB Event
The Andean cordillera behaves like multiple infrasound sources activated by the seismic wave front
Azimuth variation reflects the geomorphology of the Andes
12 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Infra-Seismic stations co-located in CanadaInfra-Seismic stations co-located in Canada
Mining activities recorded at the Canadian stations: IS10 + ULM
Lake Winnipeg
Lake Superior
IS10
200km
400km
ULM
13 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
PMCC analysis at IS10 PMCC analysis at IS10 (Canada)(Canada)
14 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Identification of the events recorded at IS10Identification of the events recorded at IS10M
ES
AB
I R
AN
GE
(40
0km
SE
)
LO
CA
L M
INE
S
Categorization of station-specific events
200k
m N
E
15 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Infrasound and meteorology: microbaroms at IS26 Infrasound and meteorology: microbaroms at IS26 (Germany)(Germany)
• Detections with similar characteristics over a long time period
• Detections produced by DFX-Libinfra, confirmed by PMCC
• No correlation with any microbaroms at Flers (France)
• Meteorology data: storm in the North Sea the same day
16 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
• Possible cause of the problems • Software related: algorithm, configuration• Data related• Case of IS26 (Germany)
Problems with data processing at the IDCProblems with data processing at the IDC
Number of daily detections
0 1-10 11-50 >50 Total
IS07 days(%) days
93%26
7%2
0%0
0%0
100%28
IS08 days(%) days
66%53
30%24
4%3
0%0
100%80
IS10 days(%) days
9%6
29%20
52%35
10%7
100%68
IS26 days(%) days
63%137
19%41
9%21
9%19
100%218
IS34 days(%) days
80%62
18%14
1%1
1%1
100%78
• Low and irregular number of detections observed for all infrasound stations processed at the IDC
17 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Problems with data processing at the IDCProblems with data processing at the IDC (Cont.)(Cont.)
• Asynchronous Testbed and Operations pipelines• Data is forwarded from ops to Testbed in near real time• Data is organized in time intervals of 30min for infra• Intervals are scheduled for processing when they contain a sufficient
amount of data– Problem 1: Intervals are processed in near real time: all data needed may
not be available for use at the time it is processed– Problem 2: Interval generation and execution is asynchronously done on
ops and Testbed
• Consequences for the detection process:• Ops and Testbed may process different data• Ops and Testbed may produce different detections
18 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Problems with data processing at the IDC - ExampleProblems with data processing at the IDC - Example
Testbed : I26DE (processing time interval [16:00 - 16:30] on October 16, 2001)
++ tuxshell 2001-10-16 16:34:10 time=1003248000.000 endtime=1003249800.000 name=I26DE get_infra_processing_interval: Requested processing interval 1003248000.000 to 1003249800.000 get_infra_processing_interval: Computed processing interval 1003247999.983 to 1003249632.033 get_infra_processing_interval: Actual waveform data interval 1003247860.885 to 1003249839.957-- tuxshell 2001-10-16 16:35:01 time=1003248000.000 endtime=1003249800.000 name=I26DE
Operations : I26DE (processing time interval [16:00 - 16:30] on October 16, 2001)
++ tuxshell 2001-10-16 16:34:26 time=1003248000.000 endtime=1003249800.000 name=I26DE get_infra_processing_interval: Requested processing interval 1003248000.000 to 1003249800.000 get_infra_processing_interval: Computed processing interval 1003247999.983 to 1003249712.033 get_infra_processing_interval: Actual waveform data interval 1003247860.885 to 1003249919.957-- tuxshell 2001-10-16 16:35:20 time=1003248000.000 endtime=1003249800.000 name=I26DE
19 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Problems with data processing at the IDC (Cont.)Problems with data processing at the IDC (Cont.)
Nominal timestampof the data frame
Start_time of subframe
Irregular Data frames (within 1 sample) 20s shift in wfdisc segments
1 sample (0.05s)
Channel subframe
20 seconds
All channel subframes of a Data Frame will have the same start time and duration to within one sample time unit. […] The nominal time field is set equal to the start time of the earliest channel subframe [...] on an interval multiple of 10 seconds and synchronized to the minute (Format and Protocol for Cont. data CD-1.1).
• Digitizing problems at IS26 causes asynchronous wfdiscs
20 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Problems with data processing at the IDC (Cont.)Problems with data processing at the IDC (Cont.)
12345 12345 12345 12345 12345 12345
5 chans 4 chans 3 chansNumber of synchronous channels:
• The number of detections is lower when Wfdisc segments are asynchronous
21 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
Problems with data processing at the IDC (Cont.)Problems with data processing at the IDC (Cont.)
Presence of multiple peaks on all channels
• origins of the peaks?
• impact on the processing?
Spectral analysis at IS26
22 International Data CentrePreparatory Commission for the Comprehensive Nuclear-Test-Ban Treaty Organization
SummarySummary
Infrasound signal analysis is a difficult exercise: collection of reference events is necessary • To increase the experience and knowledge in signal processing and
event recognition• Familiarization with the detection capability of IMS stations • First step before starting a comprehensive station tuning
Assessment of the analysis tools:• Existing automatic system (detection/association)• PMCC interactive tool for signal analysis
Problem of low number of detections produced at the IDC• Related to the software• Related to the configuration parameters (tuning)• Related to the data