TY - JOUR
T1 - Source parameter scaling for small earthquakes observed at the western Nagano 800-m-deep borehole, Central Japan
AU - Stork, Anna L.
AU - Ito, H.
PY - 2004/10/1
Y1 - 2004/10/1
N2 - Source parameters of 68 small earthquakes (0.6 < M
w < 3.0) in western Nagano, Japan, are investigated to determine the scaling of static stress drop (Δσ
s) and apparent stress (σ
a) with seismic moment (M
o). Data from the 800-m-deep borehole in the area provides clean recordings containing a wide range of frequencies. Source parameters are determined by manipulation of P and S spectra in the frequency domain using individually determined time window lengths for arrivals on each seismogram. Frequency-independent quality factors, Q, corner frequencies, f
c and the amplitude spectra levels are estimated with the best-fitting Brune (1970) ω
2 model. A frequency-dependent attenuation model, Q(f), is calculated by spectra normalization. Static stress dropΔσ
s, is self-similar for 10
10 < M
o < 10
13 N m and values range between 0.01 and 10 MPa. An F distribution test shows there is a 10% probability that σ
a calculated from Q has a constant mean value for 10
10 < M
o < 10
13 N m. Furthermore, σ
a calculated using Q(f) has a significant probability that there is a change in σ
a values with M
o. Using Q(f), 0.02 < σ
a < 2 MPa for M
o > 10
11 N m, and 0.002 < σ
a < 0.2 MPa for M
o < 10
11 N m, a narrower range than Q analysis results. Limits in recorded frequencies, variations in time window length, and source complexity are not found to significantly affect the calculation of σ
a. Therefore, the constant scaling of Δσ
s, with M
o and the nonsimilarity and breakdown in σ
a scaling could be true characteristics of small earthquakes (M
w <1.3).
AB - Source parameters of 68 small earthquakes (0.6 < M
w < 3.0) in western Nagano, Japan, are investigated to determine the scaling of static stress drop (Δσ
s) and apparent stress (σ
a) with seismic moment (M
o). Data from the 800-m-deep borehole in the area provides clean recordings containing a wide range of frequencies. Source parameters are determined by manipulation of P and S spectra in the frequency domain using individually determined time window lengths for arrivals on each seismogram. Frequency-independent quality factors, Q, corner frequencies, f
c and the amplitude spectra levels are estimated with the best-fitting Brune (1970) ω
2 model. A frequency-dependent attenuation model, Q(f), is calculated by spectra normalization. Static stress dropΔσ
s, is self-similar for 10
10 < M
o < 10
13 N m and values range between 0.01 and 10 MPa. An F distribution test shows there is a 10% probability that σ
a calculated from Q has a constant mean value for 10
10 < M
o < 10
13 N m. Furthermore, σ
a calculated using Q(f) has a significant probability that there is a change in σ
a values with M
o. Using Q(f), 0.02 < σ
a < 2 MPa for M
o > 10
11 N m, and 0.002 < σ
a < 0.2 MPa for M
o < 10
11 N m, a narrower range than Q analysis results. Limits in recorded frequencies, variations in time window length, and source complexity are not found to significantly affect the calculation of σ
a. Therefore, the constant scaling of Δσ
s, with M
o and the nonsimilarity and breakdown in σ
a scaling could be true characteristics of small earthquakes (M
w <1.3).
UR - http://www.scopus.com/inward/record.url?scp=11844255564&partnerID=8YFLogxK
U2 - 10.1785/012002214
DO - 10.1785/012002214
M3 - Article (Academic Journal)
AN - SCOPUS:11844255564
SN - 0037-1106
VL - 94
SP - 1781
EP - 1794
JO - Bulletin of the Seismological Society of America
JF - Bulletin of the Seismological Society of America
IS - 5
ER -