G E O Z E N T R U M H A N N O V E RG Z H

for arthquakes in GermanyE

ALISE - An ert and nformation ystemAl I S

Impact Quantification

Workflow

Instrumental Intensity Vulnerability

Consistency checks areused to assess the qualityof the epicenterdetermination. The resultof the spatial consistencyis shown on the left. Anumerical algorithm isused to calculate theconsistency value Cp.This value quantifies theprobability whether theepicenter locationmatches the distributionof stations contributingphases. Results of theautomatic dataprocessing with epicenterlocations in the blue areaare rejected.

EventGrouping

ConsistencyChecks and

QualityEstimation

ConsistencyChecks and

QualityEstimation

HyposatLocalization

ImpactQuantification

HyposatLocalization

Grid-SearchAlgorithm

Initial Solution

Near Real TimeWaveform Data

S1

S2

S3

S4

S5

- Signal-Definition- Dominant Frequency- Polarisation of Phases- f,k-Analysis for Array Data

...

Waveform-Processing

Information Handler

Prepare Information for furtherProcessing

Bulletin

Interactive Post-Processing andQuality Control

Large EventALERT

STA/LTADetector

S1

S2

S3

S4

S5

Phase-TypeSearch

Database

Notify byEmail / SMS

- Source Parameter- Intensity Map- Vulnerability Map- Seismicity

...

Create Webpage with

Final SolutionDatabase

Geo- andEvent Data

Jackknife PhaseSelection

German EventMap Server

ALISE is a newly developed earthquake alert and information system for Germany that consists ofthe following main components:

A seismometer network of selected stations of the German Regional Seismic Network (GRSN)and two additional new stations.

A communication system based on the CD1.0 protocol that enables near real-timetransmission of waveform data from the seismometer stations to the central recording andprocessing system.

Newly developed processing software for rapid and reliable event parameter determination.

A tool for distributing alert messages via SMS and e-mail and to display the resultsautomatically on the Web.

An interface to a newly developed map server that provides detailed information on theepicenter region and the potential damage.

Until today, two new seismic stations have been built northeast of Hannover and in centralGermany close to Bad Hersfeld. The CD1.0 communication protocol ensures immediate access tothe most recent data at any time, even after a acommunication link was disrupted.

Data processing for the determination of seismic source parameters is the most important elementof the alert system. The software makes use of improved procedures for seismic phaseidentification and phase association. Moreover, several consistency checks are included. Theapplication of these automatic methods provides reliable estimate of the source parameters.

A graphic user inter-face was developed fortesting the algorithmsof the automatic dataprocessing. This toolis perfectly suited forprocessing both, nearreal time data and fileswith selected data setsof earthquakes withwell known sourceparameters.The GUI acts as afront-end of thecontinuous detectionprocess.

Overview

Graphic User Interface

Initial Location ErrorEarthquake Test Data

Signal Definition

Multiphase GridsearchJackknife Analysis Consistency Checks

Smoothed envelopes of seismograms are used to define the signalwindows. These are needed by the waveform algorithms thatcompute varius signal parameters such as dominant frequency,polarization etc.

A multiphase-gridsearch algorithm was developed to calculate a stableinitial epicenter location which is required for subsequent signal analysis.The figure shows the computed fit between the phases an the corres-ponding grid points for an event in central Germany. The diagram belowshows the fit as a function of distance from the epicenter.

The magnitudes of selected events to testthe performance of the alert system areML>3.5 in Germany and ML> 4.0 outside.

To quantify the impact of anearthquake in Germany theinstrumental intensity (afterMSK) is determined from thesource parameters.Additionally, an estimate ofthe vulnerability is provided.This calculation takes thepopulation density intoaccount. The correspondingresults for a fictitiousmagnitude 6.0 earthquakesouth west of Cologne areshown on the right..

After determination of the final epicenter location and estimation of the earthquake’s impact, theresults are distributed via SMS and email and displayed on the Web together with additionalstatistical and geological information provided by the map server.

Phases which bias thelocalization are determinedand removed. This is doneby a Jackknife method whichperforms localizations with asubset of all phases. Fromthe change of the localizationerror the (good or bad)influence of the omittedphases is deduced. Finally, aset of good hypocentersolutions remain, shown inthe figure by black ringswhose diameters resemblethe spatial error of thesolution.

Polarization Analysis

Three component polarization analysis is applied to theseismic signals to determine the back azimuth. The quality ofthe results depends on the signal to noise ratio and on thegeological conditions below the seismometer site. Results ofpolarization analysis are shown in the two figures above with aP-phase on the left and S-phase on the right side,respectively.

Error of initial localization obtained for the test data set asa function of the gap. For events covered by the networkwith a gap < 180 deg the location difference is less than30 km.