by Dong-Hoon Sheen, Tae-Seob Kang, and Junkee...

A Local Magnitude Scale for South Korea

by Dong-Hoon Sheen, Tae-Seob Kang, and Junkee Rhie

Abstract A new local magnitude scale was developed for South Korea using seis-mograms of 269 earthquakes selected in the magnitude range 2.0–5.8 that occurred inand around the Korean Peninsula from 2001 to 2016. The peak amplitudes of syntheticWood–Anderson seismograms were measured from three components of 6327 obser-vations recorded at distances of 10–600 km by 89 broadband seismic stations in SouthKorea. The vertical peaks and geometrical means of the horizontal peaks were usedseparately for both nonparametric and parametric methods. The empirical attenuationcurve, station corrections, and magnitudes of the earthquakes were estimated simulta-neously using each method, which yielded very similar results. The resulting parametricattenuation curves are logA0 � −0:5869 logR=100 − 0:001680�R − 100� − 3 for thehorizontal component and logA0 � −0:5107 logR=100 − 0:001699�R − 100� − 3

for the vertical component, in which R is the epicentral distance in kilometers.The application of these attenuation curves to the dataset showed that therewas no trend in the magnitude residual with distance. The spatial variation ofthe station corrections generally correlated with the geological features underlyingthe seismic stations. The station corrections for the horizontal component variedmore than those for the vertical component, suggesting that geological or local siteeffects have a stronger influence on horizontal amplitudes than on vertical ones. Wefound that ML from this study correlates well with Mw determined from S-wavesource spectra.

Introduction

Local magnitude (ML), which was defined by Richter(1935), is the most widely used fundamental magnitude scalein earthquake catalogs. Although it is only an empirical es-timation of the relative sizes of earthquakes, it has improvedour understanding of the seismic characteristics and seismichazard in a region.

The original ML scale was developed for crustalearthquakes that occurred in southern California, andused the peak amplitudes of the seismic waves. Therefore,to follow Richter’s definition in other areas with differentseismotectonic environments, it is necessary to calibrate thescale considering local attenuation characteristics.

In South Korea, the Korea Meteorological Administra-tion (KMA) and Korea Institute of Geoscience and MineralResources (KIGAM) provide their own parameters, such asthe location, origin time, and magnitude, for earthquakes inand around the Korean Peninsula. However, neither instituteuses magnitude relationships calibrated specifically forSouth Korea. KMA has officially monitored earthquakesoccurring in the Korean Peninsula region since 1978 and hasused a magnitude formula proposed by Tsuboi (1954) fordetermining the magnitude (Kim and Park, 2002). After dig-ital recordings became available in 1999, the formula wasmodified (Kim and Park, 2002). KIGAM also uses a modi-

fied formula (Sheen, 2015) that was originally developed forJapan by Kanbayashi and Ichikawa (1977) and Takeu-chi (1983).

The first digital seismic network in South Korea startedin the mid-1990s and the number of stations has increasedgreatly since then. Several studies examined a local magni-tude scale for the southern Korean Peninsula with seismo-grams obtained during the first few years of operation of thedigital seismic networks (e.g., Hong et al., 2000; Kim andPark, 2002; Shin et al., 2005). Because of the low-to-moderate seismicity of the peninsula (i.e., Sheen et al.,2017), the number of events, quantity of data, and range ofearthquake magnitudes were likely insufficient for magni-tude scaling studies. Therefore, the results of these initialstudies might be incomplete for official use.

Recently, Sheen (2015) showed that the station magni-tudes of KMA and KIGAM increase with epicentraldistance. This indicates that the ML scales used by bothinstitutes do not represent the characteristics of seismicattenuation in the southern Korean Peninsula accurately.Therefore, this study developed an ML scale that representsthe attenuation properties of seismic waves in the southernKorean Peninsula using a large dataset from earthquakes thatoccurred in the peninsula region from 2001 to 2016.

2748

Bulletin of the Seismological Society of America, Vol. 108, No. 5A, pp. 2748–2755, October 2018, doi: 10.1785/0120180112

Downloaded from https://pubs.geoscienceworld.org/ssa/bssa/article-pdf/108/5A/2748/4345653/bssa-2018112.1.pdfby Seoul National Univ useron 26 September 2018

Data and Preliminary Analysis

The seismograms used in this study (see Data and Re-sources) comprise 6327 observations from 89 stations of 269earthquakes that occurred in the Korean Peninsula regionfrom 2001 to 2016. This study included only observationswith records of all three components, to use the same infor-mation to compare the attenuation characteristics and stationcorrection for the horizontal and vertical components. Theinitial dataset was based on 10,076 observations for 533earthquakes and subject to quality screening, as described inmore detail below. The seismic stations are operated byKMA, KIGAM, and the Korea Institute of Nuclear Safety(KINS) and are equipped with one of three sensor types:STS-1, STS-2 or 2.5, and CMG-3T or 3TB sensors. In total,63 of the stations used in this study belong to KMA, 22 toKIGAM, and 4 to KINS. Figure 1 shows the locations of theseismic stations and epicenters resulting in ray paths cover-ing the southern Korean Peninsula and adjacent areas.

The KMA has cataloged earthquakes greater than orequal to magnitude 2.0 occurring since 1978. We selectedonly those that occurred between 2001 and 2006 and hadmagnitudes of at least 3.5. This is because there were fewerbroadband stations operating between 2001 and 2006 and

only a few seismograms from broadband stations were avail-able for earthquakes of small magnitudes. From 2007, weconsidered all events in the KMA catalog, for which themagnitudes are greater than or equal to 2.0, but excludedaftershocks of the 2016 Mw 5.6 Gyeongju earthquakes withmagnitudes less than 3.5 (Son et al., 2018) because therewere too many small aftershocks.

After careful visual inspection of the broadband seismo-grams, clipped, cut, or noisy seismograms were identifiedand excluded. Only observations with seismograms of allthree components were selected, which resulted in 10,076observations initially. After removing the mean and lineartrend, the waveforms were filtered with a 0.5–10 Hz, six-poleButterworth band-pass filter to suppress microseismic noise,corrected from their instrument responses, and convolvedwith a revised Wood–Anderson instrument response accord-ing to Uhrhammer and Collins (1990).

The zero-to-peak amplitude was measured for eachsynthesized Wood–Anderson seismogram of the three com-ponents in a velocity window of 4:0–2:0 km=s, correspond-ing to the Sg or Lg phases. Only amplitudes greater thantwice the pre-P noise, averaged for a one-second windowof the pre-P signal, were used. Observations lacking ampli-tude measurements for all three components were againexcluded. The horizontal peak amplitudes were obtainedfrom the geometric mean of two horizontal peaks. Finally,this study used only events and stations with at least 10 ob-servations at a distance ranging from 10 to 600 km, whichresulted in 6327 observations for 269 earthquakes recordedat 89 stations. Of the data, 50% are at distances between 80and 230 km and 64% are in the magnitude range from 2.5 to3.5 magnitude units (Fig. 2).

Nonparametric Method

Richter (1935) defined a local magnitude scale with theform

EQ-TARGET;temp:intralink-;df1;313;293ML � logA − logA0 � S; �1�in which A is the peak amplitude in millimeters measuredfrom a Wood–Anderson seismogram, logA0 is an empiri-cally derived attenuation correction term depending on theepicentral distance, and S is the station correction. Theattenuation correction term converts the peak amplitude intoa measurement at an epicentral distance of 100 km andanchors an amplitude of 1 mm at a distance of 100 km tobe magnitude 3.0.

The nonparametric method for local magnitude intro-duced by Savage and Anderson (1995) was rewritten equa-tion (1) in the following form:

EQ-TARGET;temp:intralink-;df2;313;136ai logA0�Ri� � bjMj − ckSk � logAjk�R�; �2�in which the indices j and k represent events and stations,respectively; M and S are the magnitudes and stationcorrections, respectively; and R is the epicentral distance in

122°

122°

124°

124°

126°

126°

128°

128°

130°

130°

132°

132°

32° 32°

34° 34°

36° 36°

38° 38°

40° 40°

42° 42°

Magnitude

5.0

4.0

3.0

2.0

Figure 1. The distribution of the ray paths used in this study be-tween earthquakes (circles) and stations (triangles). The size of eachcircle is proportional to the earthquake magnitude obtained from theKorea Meteorological Administration (KMA) catalog. The gray linesrepresent the great circle paths of the event–station pairs. The locationof the main map is shown by the red box in the inset map.

A Local Magnitude Scale for South Korea 2749


kilometers. In this study, the attenuation correction term logA0

at distances from each node of Ri are interpolated linearly torepresent the peak amplitudes logAjk measured at arbitrarydistances R using the coefficients ai � �logRi�1 − logR�=�logRi�1 − logRi�, ai�1 � 1 − ai, and all other a � 0. Thedistance intervals for the nodes Ri can be equal or unequal.Coefficients bj and ck are also weighting factors defined asbj � 1 for event j and all other b � 0, and ck � 1 for stationk and all other c � 0, respectively. The inversion was per-formed using a least-squares procedure with the constraintsΣSk � 0, logA0�100� � −3, and logA0�Ri−1�−2 logA0�Ri�� logA0�Ri�1�� 0 (Savage and Anderson, 1995). Then, thismethod yields the solution of equation (1) as the attenuationcorrections logA0�Ri� at given distance nodes, event magni-tudes Mj, and station corrections Sk.

To consider the effect of the choice of distance interval onthe attenuation corrections, nonparametric inversion wasperformed for both regular and irregular distance intervals.Figure 3 shows the differences in the distribution of the datacorresponding to the choice of distance interval, and the

results are given in Figure 4. As a regularinterval, a distance of 50 km was used,although the shortest distance nodes hadan interval of 40 km. As expected from Fig-ure 2, most of data used for the inversionwith the regular interval were distributedfrom 50 to 300 km. Irregular intervals wereused to balance the amount of data withineach interval, and the average number ofdata was 667� 47 in each interval for dis-tances less than 400 km (see Fig. 3b).

Parametric Method

The parametric method (Bakun andJoyner, 1984) introduces parameters thatconsider geometrical spreading and theattenuation of seismic waves and rewrites

the attenuation correction − logA0 in equation (1) as:

EQ-TARGET;temp:intralink-;df3;313;509− logA0�nlog�

R100

��K�R−100��3; �3�

in which n and K are parameters for geometrical spreadingand attenuation, respectively, and R is the hypocentral orepicentral distance in kilometers.

To determine n and K from the observed data, equa-tion (3) is formulated as:

EQ-TARGET;temp:intralink-;df4;313;403n logR� KR� bjMj − ckSk � C � logAjk�R�; �4�

in which R is the epicentral distance in kilometers. C is aconstant for scaling to Richter’s definition at 100 km, whichis not formulated explicitly in equation (4), but is determinedafter inverting the observations (Ottemöller and Sargeant,2013). The summation of station corrections is constrainedas 0, which is the same as in the nonparametric method.

Result and Discussion

Figure 4 plots the attenuation curves obtained from thisstudy. The symbols and lines show that Richter’s definition isimposed by setting logA0 to −3 at 100 km. The symbols andsolid lines represent the results from the nonparametric andparametric inversions in this study, respectively. The attenu-ation corrections for the horizontal and vertical componentsare shown in red and black, respectively. For comparison, thefigure also shows the attenuation curves from several otherstudies (Hutton and Boore, 1987; Kim, 1998; Ortega andQuintanar, 2005; Miao and Langston, 2007), plotted withdashed lines. The attenuation curves from the current studywere intermediate between those from the central UnitedStates (Miao and Langston, 2007) and southern California(Hutton and Boore, 1987), but are very close to that from theBasin of Mexico (Ortega and Quintanar, 2005), which wasobtained for hypocentral distances less than about 250 km.

50

100

150

200

250

300

Num

ber

of r

ecor

ds

0 200 400 600

Distance (km)

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

Mag

nitu

de

0 5 10 15 20 25 30

Number of events

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

Mag

nitu

de

0 200 400 600

Number of records

Figure 2. Magnitude and distance distribution of the data used in this study.

0

200

400

600

800

1000

1200

Num

ber

of d

ata

100 200 300 400 500 600

Distance (km)

100 200 300 400 500 600

Distance (km)

(b)(a)

Figure 3. Histograms of the distance distribution of data corre-sponding to the distance intervals for the nonparametric method.(a) Distance distribution corresponding to regular intervals. (b) Dis-tance distribution corresponding to irregular intervals.

2750 D.-H. Sheen, T.-S. Kang, and J. Rhie


The attenuation corrections at distance nodes for thenonparametric method are represented by diamonds andcircles, which refer to the results using regular and irregularintervals, respectively (see Fig. 3). There are no significantdifferences in the attenuation corrections between regularand irregular distance intervals, which suggests that the dis-cretization used for the nonparametric method has a minimaleffect on the results of this study.

With the parametric method, the attenuation curves forthe horizontal and vertical components are represented byred and black lines, respectively, and are given byEQ-TARGET;temp:intralink-;df5;55;182

− logA0 � 0:5869 log�R=100� � 0:001680�R − 100� � 3

for horizontal; �5�andEQ-TARGET;temp:intralink-;df6;55;122

− logA0 � 0:5107 log�R=100� � 0:001699�R − 100� � 3

for vertical; �6�in which R is the epicentral distance in kilometers.

The attenuation curves obtained with the parametricmethod are very similar to the attenuation corrections at dis-tance nodes with the nonparametric method (see Fig. 4). Thedifference between the results of the two methods increasedslightly beginning at a distance of about 400 km and reachedabout 0.2 magnitude units at 600 km. However, the differ-ence in the event magnitudes between the two methods isonly 0:049� 0:012 magnitude units for the horizontal com-ponent and 0:015� 0:012 magnitude units for the verticalcomponent, as shown in Figures 5a and 5b, respectively. Thisis because the event magnitude is determined by averagingthe station magnitudes at all distance ranges. Therefore, thediscrepancy in the attenuation corrections between the non-parametric and parametric methods at distances exceeding400 km is negligible.

For both the nonparametric and parametric methods, theattenuation corrections for the horizontal component areslightly larger than those for the vertical component at dis-tances greater than 100 km, but the differences are still lessthan 0.1 magnitude units. Figure 5 compares the magnitudesof the events obtained with each method. The magnitudesfrom the horizontal component are slightly larger than thosefrom the vertical component for both inversions. The differ-ence in event magnitudes between the horizontal and verticalcomponents is 0:085� 0:041 magnitude units for the non-parametric inversion and 0:051� 0:042 magnitude units forthe parametric inversion (Fig. 5c,d).

All magnitude scaling relationships developed in thisstudy give very similar event magnitudes for the horizontaland vertical components, which imply that this study pro-vides a consistent way to measure earthquake magnitudes inthe Korean Peninsula region.

Figure 6 compares the station corrections obtained fromeach inversion. The corrections for the horizontal componentvary over a wider range than those for the vertical compo-nent. The station corrections vary between −0:54 and 0.27magnitude units for the horizontal component and −0:29 and0.26 magnitude units for the vertical component, suggestingthat local site or geological effects have a stronger influenceon the horizontal amplitudes than on the vertical ones. Thenonparametric and parametric methods yield very similarstation corrections for each component and those for thehorizontal and vertical components are generally propor-tional to each other.

Figure 7 shows the spatial distributions of the stationcorrections. The corrections are generally large and negativein the southeastern part of the peninsula and Jeju Island,which correspond to the Gyeongsang (GB) and Yeonil (YB)basins, and the Quaternary Jeju volcanic terrain (JVT), re-spectively. Kang and Shin (2006) showed that the distribu-tion of the Rayleigh-wave group velocity correlates very wellwith the geological characteristics of the southern KoreanPeninsula. Considering that the sedimentary basins (e.g., GBand YB) and volcanic and metasedimentary rocks (JVT)were represented by regions with low velocities, the stationcorrections obtained from this study could also be explained

−4.5

−4.0

−3.5

−3.0

−2.5

−2.0lo

g 10A

0

0 100 200 300 400 500 600

Epicentral distance (km)

Nonparametric (horizontal/irregular)

Nonparametric (vertical/regular)

Nonparametric (horizontal/irregular)

Nonparametric (vertical/regular)

Parametric (horizontal)

Parametric (vertical)

Central U.S. (Miao and Langston, 2007)Eastern North America (Kim, 1998)Southern California(Hutton and Boore, 1987)Basin of Mexico(Ortega and Quintanar, 2005)

Figure 4. The attenuation correction curves obtained from thisstudy. Symbols and solid lines represent the nonparametric and para-metric results, respectively. Diamonds and circles are the attenuationcorrection terms at the distance nodes of regular and irregular inter-vals for the nonparametric method. The results for the horizontal andvertical components are shown in red and black, respectively. Thedashed lines are attenuation curves from other studies.



by the correlation with the geological features underlying theseismic stations.

Figure 8 shows theML residual distributions as a functionof epicentral distance for the horizontal and vertical compo-nents. The residual measures the difference between the mag-nitude at a single station and the event magnitude obtained as

the average of the individual station magni-tudes. The station magnitudes for Figure 8were determined by the results of the para-metric inversion, that is, equations (5) and(6), and the event magnitudes determinedusing the truncated mean correspond toan iterative drop in station magnitude of0.5 magnitude units above or below thearithmetic mean of the station magnitudes.The dashed lines indicate the results of thelinear regression. These showed no particu-lar trend with respect to either distance orlogarithmic distance and imply that the at-tenuation characteristics in and around thesouthern Korean Peninsula are well repre-sented by the results. The station correc-tions help to reduce the variances of theresiduals. The variances of the residualsfor the horizontal and vertical componentsare reduced by 28% and 13%, respectively,and the standard deviations of the residualsfor both are decreased to about 0.17 mag-nitude units. The differences between theevent magnitudes with and without stationcorrections are 0.026 and 0.0091 magni-tude units for the horizontal and verticalcomponents, respectively, which shows thatthe introduction of station corrections haslittle effect on the determination of theevent magnitude. However, this would beimportant for small earthquakes with fewobservations (e.g., Ottemöller and Sar-geant, 2013) and, especially, for the deter-mination from the horizontal component.

Figure 9 compares the event magni-tudes obtained from this study with Mw

of Rhee and Sheen (2016). Rhee andSheen (2016) determined Mw of an earth-quake from the vector sum of three com-ponents of S-wave source spectra. Rheeand Sheen (2016) considered only eventsbetween 2001 and 2014; as a result, 170earthquakes are shown in the figure. Theorthogonal linear regression based onthe principal component analysis givesthe following equations for the ML–Mw

relationship:

EQ-TARGET;temp:intralink-;df7;313;154ML � 1:076Mw − 0:4014 for horizontal; �7�

and

EQ-TARGET;temp:intralink-;df8;313;124ML � 1:086Mw − 0:4772 for vertical; �8�

with standard deviations of 0.1051 and 0.1049 magnitudeunits for the horizontal and vertical components, respec-

−0.6

−0.4

−0.2

0.0

0.2

0.4

Sta

tion

corr

ectio

n P

−0.6 −0.4 −0.2 0.0 0.2 0.4

Station correctionNP

HorizontalVertical

−0.6

−0.4

−0.2

0.0

0.2

0.4

Sta

tion

corr

ectio

n ver

−0.6 −0.4 −0.2 0.0 0.2 0.4

Station correctionhor

NonparametricParametric

(b)(a)

Figure 6. Comparison of the station corrections. (a) Station corrections from theparametric versus nonparametric methods. (b) Station corrections for the vertical versushorizontal components.

2

3

4

5

6P

aram

etric

ML

2 3 4 5 6

Nonparametric ML

Horizontal

2

3

4

5

6

Par

amet

ric M

L

2 3 4 5 6

Nonparametric ML

Vertical

2

3

4

5

6

Ver

tical

ML

2 3 4 5 6

Horizontal ML

Nonparametric

2

3

4

5

6V

ertic

al M

L

2 3 4 5 6

Horizontal ML

Parametric

(a) (b)

(c) (d)

Figure 5. Comparison ofML estimates obtained from this study. (a,b) Nonparamet-ric versus parametric methods for the horizontal and vertical components, respectively.(c,d) Horizontal versus vertical components for the nonparametric and parametric meth-ods, respectively.



tively. ML from the current study correlates well with Mw

determined from S-wave source spectra. However, ourML–Mw relationships differ slightly from the relation inHanks and Kanamori (1979), and our ML estimation tendsto underestimate the size of small earthquakes. Other studies(e.g., Hanks and Boore, 1984; Ben-Zion and Zhu, 2002;Bindi et al., 2005; Miao and Langston, 2007; Edwards et al.,2010; Ross et al., 2016; Deichmann, 2017) reported similarobservations for small events, and Edwards et al. (2010) ex-plained this by the interaction of attenuation, the stress drop,and the Wood–Anderson filter.

Figure 10 shows the relationships between the horizon-tal and vertical ML values from the current study and thosefrom the KMA catalog. The KMA determined ML from thelargest horizontal peak amplitude, however we determined itfrom the geometric mean of the horizontal peaks. The rela-tionships were also obtained using orthogonal linear regres-sion, given by

EQ-TARGET;temp:intralink-;df9;55;166ML � 0:9187MKMAL � 0:3906 for horizontal; �9�

and

EQ-TARGET;temp:intralink-;df10;55;120ML � 0:9234MKMAL � 0:3262 for vertical; �10�

with standard deviations of 0.1932 and 0.1980 magnitudeunits for the horizontal and vertical components, respec-

tively. In comparison with Figure 9, the correlation is ratherweak, especially for small earthquakes, which implies thatML from our study correlates better withMw than the catalogof the KMA. We also found that KMA tends to under-estimate the size of small earthquakes with greater uncer-tainty, which agrees quite well with what we expect fromSheen (2015).

Conclusion

In this study, we developed a local magnitude scale forSouth Korea using 6327 horizontal and vertical peak ampli-tudes from 269 earthquakes in the magnitude range 2.0–5.8that occurred in and around the Korean Peninsula from 2001to 2016. Seismograms were recorded at 89 broadband seis-mic stations within distances of 10–600 km. The zero-to-peak amplitudes were measured from three components ofsynthetic Wood–Anderson seismograms and the geometricmean of two horizontal peaks was used for the horizontalpeak amplitude.

We used nonparametric and parametric methods toestimate the attenuation curves, station corrections, and mag-nitudes of earthquakes; both approaches yielded very similarresults. The magnitude residuals between the individualstation magnitudes and the event magnitudes that were mea-sured with attenuation curves in this study showed no trendswith distance for either component, which implies that the at-

124(a) (b)

° 126° 128° 130°

34°

36°

38°

PBOB

IBGM

OFB

YM

GB

YB

JVT

Number ofdata

(Horizontal)200

1005010

124° 126° 128° 130°

34°

36°

38°

Number ofdata

(Vertical)200

1005010

−0.3 −0.2 −0.1 0.0 0.1 0.2 0.3

Station correction

Figure 7. Spatial distribution of the station corrections for the (a) horizontal and (b) vertical components. Both were obtained with theparametric method. The color and size of the triangles represent the scale of the correction and the amount of data used for each station,respectively. Tectonic boundaries are denoted by thick solid lines (PB, Pyeongnam basin; OB, Ongjin basin; IB, Imjingang basin; GM,Gyeonggi massif; OFB, Okcheon fold belt; YM, Yeongnam massif; GB, Gyeongsang basin; YB, Yeonil basin; JVT, Jeju volcanic terrain).



tenuation characteristics around the southern KoreanPeninsula were well represented in this study. The station cor-rections for the horizontal and vertical components are gener-ally proportional to each other. They varied between −0:54and 0.27 magnitude units for the horizontal component and−0:29 and 0.26 magnitude units for the vertical component.However, it was found that the introduction of station correc-tions had little effect on the determination of event magnitude.

The event magnitudes determined in this study were com-pared with Mw from S-wave source spectra and ML in theKMA catalog. There was a good correlation between ML

in this study and Mw. The size of small events in theKMA catalog appeared to be underestimated with large uncer-tainty, probably because of a poorly constrained attenuationcurve. Therefore, we recommend that the local magnitudescale from this study be used to determine the local magni-

−1.6

−1.2

−0.8

−0.4

0.0

0.4

0.8

1.2

1.6(a)

(b)

ML

resi

dual

(ho

rizon

tal) without correction

with correction

> 200 km100 ~ 200< 100 km

−1.6

−1.2

−0.8

−0.4

0.0

0.4

0.8

1.2

1.6

ML

resi

dual

(ve

rtic

al)

0 100 200 300 400 500 600

Epicentral distance (km)

without correctionwith correction

> 200 km100 ~ 200< 100 km

10 100

Epicentral distance (km)0 500 1000 1500

Number of data

Figure 8. Magnitude residual distribution as a function of epicentral distance for the (a) horizontal and (b) vertical components. Plussymbols and circles represent the results with and without station correction, respectively. The colors of the histograms correspond to those ofthe plus symbols, which divide the distance range of the observations. The red and black dashed lines were obtained from linear regressionsof the results with the corrections for linear and logarithmic distance scales, respectively.

2

3

4

5

6

Hor

izon

tal M

L

2 3 4 5 6

MW

2

3

4

5

6

Ver

tical

ML

2 3 4 5 6

MW

Figure 9. Comparison of the magnitudes betweenML from thisstudy and Mw from Rhee and Sheen (2016). The orthogonal linearregression results are plotted with red solid lines, and red dashedlines indicate one standard deviation.

2

3

4

5

6

Hor

izon

tal M

L

2 3 4 5 6

KMA ML

2

3

4

5

6

Ver

tical

ML

2 3 4 5 6

KMA ML

Figure 10. Comparison of the magnitudes between ML fromthis study and ML from the KMA catalog. The orthogonal linearregression results are plotted with red solid lines, and red dashedlines indicate one standard deviation.



tudes of earthquakes occurring in and around the southernKorean Peninsula, and that the catalog for South Korea berecomputed using the results of this study.

Data and Resources

The earthquake information was obtained from the KoreaMeteorological Administration (KMA) using http://www.kma.go.kr/weather/earthquake_volcano/domesticlist.jsp (last ac-cessed February 2017). Seismic records were acquired fromKMA and Korea Institute of Geoscience and Mineral Resour-ces (KIGAM). All figures were generated using GenericMapping Tools (Wessel et al., 2013). The geotectonic lineswere obtained from KIGAM (https://mgeo.kigam.re.kr/, lastaccessed May 2016).

Acknowledgments

The authors thank Won-Young Kim for valuable discussion whilepreparing this article. This work was funded by the Korea MeteorologicalAdministration Research and Development Program under Grant NumbersKMIPA 2016-3010 and KMIPA 2017-4030.

References

Bakun, W. H., and W. B. Joyner (1984). The ML scale in central California,Bull. Seismol. Soc. Am. 74, 1827–1843.

Ben-Zion, Y., and L. Zhu (2002). Potency-magnitude scaling relations forsouthern California earthquakes with 1:0 < ML < 7:0, Geophys. J.Int. 148, F1–F5.

Bindi, D., D. Spallarossa, C. Eva, and M. Cattaneo (2005). Local andduration magnitudes in northwestern Italy, and seismic moment versusmagnitude relationships, Bull. Seismol. Soc. Am. 95, 592–604.

Deichmann, N. (2017). Theoretical basis for the observed break in ML=Mw

scaling between small and large earthquakes, Bull. Seismol. Soc. Am.107, 505–520.

Edwards, B., B. Allmann, D. Fäh, and J. Clinton (2010). Automatic com-putation of moment magnitudes for small earthquakes and the scalingof local to moment magnitude, Geophys. J. Int. 183, 407–420.

Hanks, T. C., and D. M. Boore (1984). Moment-magnitude relations intheory and practice, J. Geophys. Res. 89, 6229–6235.

Hanks, T. C., and H. Kanamori (1979). A moment magnitude scale, J. Geo-phys. Res. 84, no. B5, 2348–2350.

Hong, T., C. Baag, and J. Shin (2000). The ML scale in southern Korea, J.Geol. Soc. Korea 36, 545–558 (in Korean with English abstract).

Hutton, L. K., and D. M. Boore (1987). TheML scale in southern California,Bull. Seismol. Soc. Am. 77, 2074–2094.

Kanbayashi, Y., and M. Ichikawa (1977). A method for determining magni-tude of shallow earthquakes occurring in and near Japan, Q. J. Seismol.41, 56–61 (in Japanese with English abstract).

Kang, T.-S., and J. S. Shin (2006). Surface-wave tomography from ambientseismic noise of accelerograph networks in southern Korea, Geophys.Res. Lett. 33, L17303, doi: 10.1029/2006GL027044.

Kim, S. K., and M. A. Park (2002). The local magnitude scale in the KoreanPeninsula, J. Geol. Soc. Korea 38, 217–235 (in Korean with Englishabstract).

Kim, W.-Y. (1998). The ML scale in eastern North America, Bull. Seismol.Soc. Am. 88, 935–951.

Miao, Q., and C. A. Langston (2007). Empirical distance attenuation and thelocal-magnitude scale for the Central United States, Bull. Seismol. Soc.Am. 97, 2137–2151.

Ortega, R., and L. Quintanar (2005). A study of the local magnitude scale inthe Basin of Mexico: Mutually consistent estimates of logA0 andground-motion scaling, Bull. Seismol. Soc. Am. 95, 605–613.

Ottemöller, L., and S. Sargeant (2013). A local magnitude scale ML for theUnited Kingdom, Bull. Seismol. Soc. Am. 103, 2884–2893.

Rhee, H.-M., and D.-H. Sheen (2016). Lateral variation in the source param-eters of earthquakes in the Korean Peninsula, Bull. Seismol. Soc. Am.106, 2266–2274.

Richter, C. F. (1935). An instrumental earthquake magnitude scale, Bull.Seismol. Soc. Am. 25, 1–32.

Ross, Z. E., Y. Ben-Zion, M. C. White, and F. L. Vernon (2016). Analysisof earthquake body wave spectra for potency and magnitude values:Implications for magnitude scaling relations, Geophys. J. Int. 207,1158–1164.

Savage, M. K., and J. G. Anderson (1995). A local-magnitude scale for theWestern Great Basin–Eastern Sierra Nevada from synthetic Wood–Anderson seismograms, Bull. Seismol. Soc. Am. 85, 1236–1243.

Sheen, D.-H. (2015). Comparison of local magnitude scales in SouthKorea, J. Geol. Soc. Korea 51, 415–424 (in Korean with English ab-stract).

Sheen, D.-H., J.-H. Park, H.-C. Chi, E.-H. Hwang, I.-S. Lim, Y. J. Seong,and J. Pak (2017). The first stage of an earthquake early warningsystem in South Korea, Seismol. Res. Lett. 88, 1491–1498.

Shin, J., H. Chi, and C.-S. Cho (2005). Review on ML scales in southernKorea, J. Korean Geophys. Soc. 8, 207–209 (in Korean with Englishabstract).

Son, M., C. Cho, J. Shin, H.-M. Rhee, and D.-H. Sheen (2018). Spatiotem-poral distribution of events during the first three months of the 2016Gyeongju, Korea, earthquake sequence, Bull. Seismol. Soc. Am. 108,210–217.

Takeuchi, H. (1983). Magnitude determination of small shallow earthquakeswith JMA electromagnetic seismograph model 76, Q. J. Seismol. 47,112–116 (in Japanese with English abstract).

Tsuboi, C. (1954). Determination of Gutenberg–Richter’s magnitude ofearthquakes occurring in and near Japan, Zisin 2, 7, 185–193 (in Jap-anese with English abstract).

Uhrhammer, R. A., and E. R. Collins (1990). Synthesis of Wood–Andersonseismograms from broadband digital records, Bull. Seismol. Soc. Am.80, 702–716.

Wessel, P., W. H. F. Smith, R. Scharroo, J. F. Luis, and F. Wobbe (2013).Generic mapping tools: Improved version released, Eos Trans. AGU94, 409–410.

Department of Geological EnvironmentFaculty of Earth Systems and Environmental SciencesChonnam National UniversityGwangju 61186, South [email protected]

(D.-H.S.)

Pukyong National UniversityBusan 48513, South Korea

(T.-S.K.)

Seoul National UniversitySeoul 08826, South Korea

(J.R.)

Manuscript received 16 April 2018;Published Online 31 July 2018



http://www.kma.go.kr/weather/earthquake_volcano/domesticlist.jsp





https://mgeo.kigam.re.kr/

http://dx.doi.org/10.1029/2006GL027044

by Dong-Hoon Sheen, Tae-Seob Kang, and Junkee...

Documents

Transcript of by Dong-Hoon Sheen, Tae-Seob Kang, and Junkee...