DISCRIMINATING SMALL EARTHQUAKES FROM QUARRY BLASTS USING PEAK AMPLITUDE RATIO - Vmax/Hmax MA, S., EATON, D., & DINEVA, S. Department of Earth Sciences.

DISCRIMINATING SMALL EARTHQUAKES FROM QUARRY BLASTS USING PEAK AMPLITUDE RATIO - Vmax/Hmax

MA, S., EATON, D., & DINEVA, S. Department of Earth Sciences

The University of Western Ontario London, Ontario, Canada, N6A 5B7

The key differences between blasts and earthquakes:

(1) Focal mechanism – the type of source for blasts are explosive; -- the type of source for earthquakes are dislocation(2) Focal depths – the depths of blasts are on the surface or shallow;

-- the depths of earthquakes are deeper (in southern Ontario

earthquakes occur at depth ~ 3 km or deeper).(3) Source time function – in a blast generally there are several

explosions; one rupture in a small earthquake generally.

These basic differences can principally be used to distinguish blasts from small earthquakes.

Many scientists have studied blasts, for example: The following 4 slides produced by D. Craig Pearson, Brian W. Stump & David P. Anderson can help us clarify some concepts.

Physical Constraints on Mining Explosions

D. Craig Pearson Los Alamos National Laboratory

Brian W. Stump Southern Methodist University, Los Alamos National Laboratory

David P. Anderson Southern Methodist University

Prepared for Seismological Research Letters Last update 15 May 1996

[email protected]

Blast model by D. Craig Pearson, Brian W. Stump & David P. Anderson

Produced by D. Craig Pearson, Brian W. Stump & David P. Anderson

Produced by D. Craig Pearson, Brian W. Stump & David P. Anderson

Hourly distribution of seismic events in southern Ontario & western Quebec

0

200

400

600

800

1000

1200

Local time

Nu

mb

er

of

all

ev

en

ts

0

10

20

30

40

50

60

Nu

mb

er

of

kn

ow

n

bla

sts

/ea

rth

qu

ak

es

All events Blasts Earthquakes

There are many seismic events in western Quebec and southern Ontario. We need to separate earthquakes and blasts for seismic related work & projects -- only around 10% of the 8000 events recorded from 2002 to 2005 have basic parameters. The red curve shows the distribution of the known blasts; the green curve shows the distribution of earthquakes alone.

Epicentre distribution for earthquakes from 2000-10-09 to 2005-07-23; blast sites:St. Mary’s Cement site, ACTON sites A1 & A2; stations WLVO & ACTO

Waveforms generated by a blast at St. Mary’s Cement; recorded at POLARIS station WLVO; station distance ~ 24 km.

Waveforms generated by a blast at ACTON site; recorded at POLARISstation ACTO; station distance ~ 13 km

Waveforms generated by a blast at St. Mary’s Cement; recorded at local station WLVO; station distance ~ 24 km.

Waveforms generated by a small earthquake (mN~1.1); recorded at POLARIS station STCO; the distance ~18 km.

We are trying to use the ratio information to separate earthquakes from blasts.

We see that in general, ground motion produced by blasts is dominated by horizontal motion, whereas local earthquakes produce a relatively stronger vertical ground motion. The vertical ground motion from earthquakes arises from non-horizontal P waves, which arrive sooner than the strong surface-wave ground motion from blasts. Based on these observations, we are motivated to seek a simple blast discriminator based on the ratio of peak vertical to horizontal amplitude. We see the amplitudes are weaker along vertical waveforms generated by blasts at ACTON and St. Mary’s sites, so we are trying to identify blasts from small earthquakes using the ratio between the maximum amplitudes along vertical and horizontal components.

Some people have used ratio information to study earthquakes and explosions, for example:

Pg/Lg ratio has been used by W-Y Kim et al. (BSSA, 1997, 87, 569-588).

Here we use Vertical(max)/Horizontal(max) ratio. We can use computer search the two factors automatically in a very short time. We made tests with waveforms generated by blasts at ACTON site and

St. Mary’s site, as well as some small earthquakes.

In the formula: Vmax=the maximum amplitude along vertical record;

The other part is the maximum amplitude along horizontal component

ra tioV

E W N S

m ax

m ax( )2 2

Ratios: the dark blue line was drawn with the ratios of blasts at ACTON sites; the yellow line was drawn with the ratios of blasts at St. Mary’s Cement site; the purple red line was drawn with the ratios of small earthquakes in southern Ontario (43-44N; 78-80W; 2000-10-09 to

2005-07-23).

Quakes

ACTONSt Mary’s

V/H

f xx

( , , ) ex p( )

1

2 2

2

2

In the formula µ is the average; σ is the standard deviation.

For the 8 blasts at ACTON sites, µ = 0.188584; σ = 0.05335464

For the 16 blasts at St. Mary’s Cement site,µ = 0.29861623; σ = 0.06166254

For the 15 small earthquakes in the southern Ontario region,µ = 0.452439; σ = 0.13809767

Normal distributions: the dark blue line was drawn with the ratios of blasts at ACTON sites; the yellow line was drawn with the ratios of blasts at St. Mary’s Cement site; the purple red line was drawn with the ratios of small earthquakes in southern Ontario (43-

44N; 78-80W).

ACTON

St Mary’s

Quakes

Waveforms generated by a blast at St. Mary’s Cement; recorded at local station WLVO. The peak time positions along vertical and horizontal components are generally different!

Ratio_time difference: The dark blue diamond were drawn with the ratios of blasts at ACTON sites; the yellow triangles were drawn with the ratios of blasts at St. Mary’s Cement site; the purple red squares were drawn with the ratios of small earthquakes in southern Ontario (43-44N; 78-80W).

Quakes

ACTON

St Mary’s

Normal distributions: the dark blue line was drawn with the time differences between Vmax and Hmax from blasts at ACTON sites; the yellow line was drawn with the time differences between Vmax and Hmax from blasts at St. Mary’s Cement site; the purple red line was drawn with the time differences between Vmax and Hmax from small earthquakes in southern Ontario (43-44N; 78-80W).

ACTON

St Mary’s

Quakes

Quakes

ACTONSt Mary’s

V/H

Conclusion: The ratio averages for blasts in the 3 sites are small; for the studied earthquakes the average is large.

Synthetics generated using explosive source

The 3 component records at GAC, generated by a blast in western Quebec (2002-06-26)

The 3 component records at OTT, generated by the same blast at western Quebec(2002-06-26).(?)