SG Seismology VO
From EGEE-see WIki
Contents |
SEE-GRID-SCI Seismology VO
Core services
| Service | Primary location | Backup location |
| VOMS | voms.ulakbim.gov.tr | |
| WMS | wms.ulakbim.gov.tr | grid-wms.ii.edu.mk |
| BDII | bdii.ulakbim.gov.tr | grid-bdii.ii.edu.mk |
| LFC | lfc.ulakbim.gov.tr | grid-lfc.ii.edu.mk |
| FTS | fts.ulakbim.gov.tr | |
| AMGA | amga.ulakbim.gov.tr | |
| MyProxy | myproxy.ulakbim.gov.tr |
Applications
| Acronym | Application name | Application status | Interaction type | Development time | Main developer(s) |
| SRA | Seismic Risk Assesment | Development Mode | Batch | ? | Middle East Technical University (TR) |
| SDS | Seismic Data Server | Development Mode | Batch | ? | Bogazici University (TR) |
| ELF | Earthquake Location Finding | Development mode | Batch | ? | Bogazici University (TR) |
| FPS | Fault Plane Solution | Developing mode | Batch | ? | Bogazici University (TR) |
| NMMC3D | Numerical Modeling of Mantle Convection | Development Mode | Batch | ? | Geodetic and Geophysical Research Institute Seismological Observatory (HU) |
| MDSSP-WA | Massive Digital Seismological Signal Processing with the Wavelet Analysis | Development mode | Batch | ? | University of Ss. Cyril and Methodius (MK) |
Members and organization
| Seismology VO | Name | Institution | |
| VO Leader | Can Özturan | Bogazici University (TR) | ozturaca at boun.edu.tr |
| Scientific Representative | Ljupco Jordanovski | UKIM (MK) | ljordanovski at gmail.com |
| Technical Representative | Can Özturan | Bogazici University (TR) | ozturaca at boun.edu.tr |
Seismology VO Guide
Introduction
Seismology is the study of seismic phenomena that appear in the form of earthquakes and vibrations of the Earth either due to natural or artificial causes. Scientists in the seismology field need both computational resources to solve equations that arise in the mathematical modeling of seismic phenomena as well as storage resources in order to store and access historical earthquake information and massive seismic data that are collected either in continuous or triggered fashion from several geographically distributed sensors.
Before the evolution of grid infrastructures, the web was the main medium of delivery for seismic data. Seismic data providers would place their data on web servers from which scientists could either query or download the data that were of interest to them. This arrangement, however, had several shortcomings. Firstly, data and computational resources are basically decoupled. It is the duty of each scientist to manually download and manage all the data himself. Automation of this process may require the writing of various scripts which may be quite difficult for a user. Secondly, retrieving massive data over the web can be too slow and hence not practical. There have been efforts also to make seismic data available by web services which will help in automation, but again this requires users to learn web services programming and to link such services with their applications. Grids solve these problems by offering a platform where computational storage resources and other miscellaneous resources are available all connected by high speed networks. A grid user can write a program, run it on the grid where it can access distributed data just like accessing local data with the help of middleware and other tools.
The aim of the SEEGRID-SCI Seismology Virtual Organization is to bring to the griplatform seismology data and applications from the Southeastern European countries. Even though there are other organizations and projects working on seismology work, in SEEGRID-SCI redundant and duplicate effort will be avoided and complementary work will be contributed by differentiating by the grid platform that is being worked on, the performance aspects (high performance computing, high performance access to massive data) and the regional aspects (Southeastern European countries). The following figure (Figure 1) summarizes the Seismology VO platform that will be developed in the SEEGRID-SCI project:
Figure 1. Seismology VO platform
To realise the aims of the Seismology VO, the following will be implemented:
- 1. Distributed storage of seismic data from different partner countries,
- 2. Logical organization, indexing and update of distributed seismic data,
- 3. Programming tools that will provide easy access to seismic data,
- 4. Gridification of various seismology applications.
In Seismology VO, as data, (i) official lists of earthquakes and stations, and (ii) massive seismic waveform data from various Southeastern European countries are planned to be collected and served. As applications, (i) Seismic Risk Assesment (SRA), (ii) Numerical Modelling of Mantle Convection (NMMC3D), (iii) Fault Plane Solution (FPS), (iv) Automatic Earthquake Center Location (ELF), and (v) Massive Digital Seismological Signal Processing with the Wavelet Analysis (MDSSP-WA) will be gridified. To collect and serve data to the applications, a seismic data server application (SDS) is also being developed. As part of this application, programming tools in the form of iterators will be provided to the users to let them to easily go through the available station, earthquake and seismic waveform data files.
Currently, seismic data from five countries, namely Albania, Armenia, FYR of Macedonia, Moldova, and Turkey is planned to be served. Statistics about the numbers of stations in these countries are shown in Table 1. Other countries will also be able to join later as data providers.
| Country | No. of Stations |
|---|---|
| Albania | 13 |
| Armania | 326 |
| FYR of Macedonia | 6 |
| Moldova | 5 |
| Turkey | 114 (Boğaziçi-Kandilli)
+ 200 (Gen. Dir. of Disaster Affairs, Earthquake Res. Inst.) |
In the following sections, we will describe how SDS organizes the seismic data and what to set up in order to use the scripts provided by SDS to store seismic data. In Section 2, the organization of seismic data will be presented. In Section 3, necessary set-up in your home directory will be given. In Section 3, the script to enter earthquake data will be explained. In Section 4, the script to enter waveform data will be explained. In Section 5, the script to enter station data will be explained. Finally, in Section 6, we briefly present the SDS C++ iterators that is currently being programmed (and not ready) and that will let users access the data in AMGA tables.
Seismic Data Storage and Organization
Seismic waveform data and station pole zero files provided by data sources are stored in storage elements and registered in LFC. The following is the logical file hierarchy used by SDS. Ulakbim’s lfc.ulakbim.gov.tr LFC catalog is used for this purpose (Figure 2). Note that SDS scripts are also stored here and can be copied from here to your home in order to use them.
/grid/grid/seismo.see-grid-sci.eu/data/
albania/
Waveformdata_Years
Station_PoleZeroFiles/
armenia/
Waveformdata_Years
Station_PoleZeroFiles/
macedonia/
Waveformdata_Years
Station_PoleZeroFiles/
moldova/
Waveformdata_Years
Station_PoleZeroFiles/
turkey/
kandilli/
Waveformdata_Years
Station_PoleZeroFiles/
afetdad/
Waveformdata_Years
Station_PoleZeroFiles/
/tools/
scripts/
insert_waveformdata.sh
insert_earthquake.sh
Earthquake and station data are stored in AMGA tables. Metadata about the seismic waveform files are also stored in AMGA tables. The seismic waveform AMGA table acts as an index which keeps track of every single seismic waveform stored in the seismology VO. As a result, the user does not have to go through the slow and cumbersome traversal of the LFC in order to see what files are available in the seismology VO. Queries to AMGA, can give these faster. C++ iterators that will be provided by SDS later, can do this even at a higher level that can eliminate the need to learn AMGA queries.
.mdclient.config file
host = amga.ulakbim.gov.tr Port = 8822 # User settings Login = seismo PermissionMask = rwx GroupMask = r-x Home = / # Security options UseSSL = require # Values: require, try, no. If off, all options below are ignored AuthenticateWithCertificate = 1 # Use certificate to authenticate CertFile=$HOME/.globus/usercert.pem KeyFile=$HOME /.globus/userkey.pem UseGridProxy = 1 IgnoreCertificateNameMismatch = 1 # If server certificates are verified, CA certificates need to be loaded: TrustedCertDir = /etc/grid-security/certificates
Note that $HOME is the pathname of your home. Also note that login name is seismo.
Script to enter earthquake data (insert_earthquake.sh)
This scripts inserts earthquake information into an AMGA earthquakes table. You should get a seismo.see-grid-sci.eu proxy (i.e. issue the command voms-proxy-init --voms seismo.see-grid-sci.eu) before you can use this script. The earthquake information must be entered into a file in which earthquake details appears on each line. If Latitude gives the location of a place on Earth North, Latitude is positive. If Latitude gives the location of a place on Earth South, Latitude is negative. If Longitude is the east geographic coordinate measurement, Longitude is positive. If Longitude is the west geographic coordinate measurement, Longitude is negative. Each line which reports an earthquake has the following format (with fields separated by a semi-colon).
YYYYMMDDTHHMMSS;Latitude;Longitude;Depth;MD;ML;MS;MB;MW;Intensity;Country;Region
An example file showing earthquake information is shown below:
20080911T235010;38.3853;27.042;209;3.1;3.2;3.3;3.4;3.5;4.1;TR;Tecirli (Osmaniye) 20080710T225009;40.4753;39.051;310;4.1;4.2;4.3;4.4;4.5;5.1;TUR;Bozcaada (Canakkale) 20081009T154220;25.6253;27.042;410;3.2;3.4;3.9;3.2;3.7;3.3;TR;Otlukbeli (Erzincan)
Note you can have multiple lines in the file and hence report multiple earthquakes in the same file. Also note that YYYYMMDDTHHMMSS is the ISO8601 date-time format. It is possible that some fields on each line can be left empty as the following example shows.
20090911T235010;38.3853;27.042;209;3.1;3.2;3.3;3.4;3.5;;TR;Tecirli (Osmaniye)
In the above example, the intensity value has been omitted. Suppose that the name of the earthquake file is earthquakesfile. Then the command to run the script is as follows:
./insert_earthquake.sh earthquakefile
Script to enter waveform data (insert_waveformdata.sh)
In order to use this script, you should get proxy for seismo.see-grid-sci.eu. This script can then be used to (i) save seismic waveform data file to the storage element (ii) register the file in LFC and (iii) add metadata about the file to an AMGA table. The name of the file containing seismic waveform data must have the following format:
StartDateTime-EndDateTime-Country-Organization-Station-Sensor.FileFormat
Stations have three types of sensor. These are BHE, BHN, and BHZ. Country codes can be two characters or three characters ( TR or TUR ) .
Wavefrom data must not cross monthly boundaries.
Note that this script can insert multiple files. The files are expected to be located in a directory. For example, suppose we have the following waveform files in a directory:
20080911T235010-20080912T213020-TR-KAN-IST-BHE.SAC 20090910T225010-20090912T215020-AL--IST-BHN.SAC
insert_waveformdata.sh command expects the waveform files directory and the storage element on which it will be stored. An example command is given below:
./insert_waveformdata.sh /home/lufer/waveformdata_directory/ se.ulakbim.gov.tr
Entry of Station Data
Entry of information about stations is done manually by SDS developers. Data providers are supposed to provide 3 types of information:
- (i) General information about stations (see Table 2 as an example).
- (ii) Detailed information about sensors in the stations (see Table 3 as an example).
- (iii) Response file ( pole zero) of stations (see the ADVT.PZ.BHE example file)
| No | Country | Stn | Lat (deg) | Lon (deg) | Elev (m) | Serial (sensor-dig) | Install. DateTime | T (Period) | Sensor | Digitizer | Acquisition Software |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | TR | BLCB | 38.3853 | 27.042 | 50 | T3727-D0935 | 10/4/2004 | 360 | CMG-3TD | CMG-DM24 | Scream |
| 2 | TR | CTYL | 41.476 | 28.2897 | 77 | T3X32-A652 | 9/29/2007 | 360 | CMG-3TD | CMG-DM25 | Scream |
| ID | Country | STN | BuiltDateTime | CMP | NrmFac | Diff | VelOutput | Conv | vSens | Sens | Truns | Const | nPoles | nZeros | Pole1 | Pole2 | Pole3 | Pole4 | Pole5 | Pole6 | Pole7 | Pole8 | Zero1 | Zero2 | Zero3 | Zero4 | Zero5 | Zero6 | Zero7 | Zero8 | Instrument | Sample |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | TR | ADVT | 20060519 | BHZ | 2304000 | 2 | 2982 | 3.16E-06 | 1.06E-09 | 5.30E-10 | 3.159 | 4.35E+15 | 5 | 2 | -5.89e-03+j5.89e-03 | -5.89e-03-j5.89e-03 | -180j0 | -160j0 | -80j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | CMG-3ESP/T3Y19 | 50 |
| 2 | TR | ADVT | 20060519 | BHN | 2304000 | 2 | 2987 | 3.17E-06 | 1.06E-09 | 5.31E-10 | 3.174 | 4.34E+15 | 5 | 2 | -5.89e-03+j5.89e-03 | -5.89e-03-j5.89e-03 | -180j0 | -160j0 | -80j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | 0j0 | CMG-3ESP/T3Y19 | 50 |
Response file ( pole zero) of stations
Example :
File name : ADVT.PZ.BHE
File content:
ZEROS 2 0 0 0 0 POLES 5 -0.00589 0.00589 -0.00589 -0.00589 -180 0 -160 0 -80 0 CONSTANT 4.15E+15
