基于概率密度函数表示的地震计自噪声模型

doi:10.3969/j.issn.0253-4967.2023.05.009

摘要/Abstract

摘要：

不同型号地震计的自噪声水平直接影响了地震观测数据的质量, 并进一步限制了利用地震数据解决地球科学问题的能力。长久以来, 由于受到观测条件的限制, 准确地测量和比较不同型号地震计的自噪声水平颇具挑战。文中利用马陵山地震台4个月的连续地震波形, 基于概率密度函数的表示方法计算了7个型号地震计的自噪声功率谱密度曲线。对于超宽频带地震计CMG-3T-360, 在微震频带(0.1~1Hz), 水平方向的自噪声明显高于垂直方向, 这可能是地震计水平方向相对方位未对齐导致的计算偏差; 在低频段(<0.03Hz), 水平方向显著偏高的自噪声可能源于大气压的变化。短周期地震计JS-S02的自噪声水平在频率为0.15~7Hz时低于全球新低噪声模型(NLNM)。宽频带地震计TDV-60B和甚宽频带地震计TDV-120VB在垂直方向的自噪声水平基本一致。宽频带地震计JS-60和甚宽频带地震计JS-120的自噪声水平在微震频段接近或低于CMG-3T-360。当频率为0.008~0.08Hz时, JS-120水平向NS通道的自噪声水平高于CMG-3T-360功率谱密度68%的置信区间。

关键词: 地震计, 自噪声, 功率谱密度, 概率密度函数

Abstract:

The self-noise levels of different seismometer models directly affect the quality of seismic observations and further limit earth science research based on seismic records. We conducted a test of the self-noise level of seven types of seismometers at the Malingshan seismic station, in which the instrument types included short-period, broadband, very-broadband and ultra-broadband. Three seismometers of each type were set up, and the observation period was from November 22, 2018, to March 26, 2019. In this paper, based on continuous seismic waveforms from seven models of seismometers, the self-noise power spectral density(PSD)of the seismometers was calculated by using the three sensor correlation analysis method, and the probability density distribution of the self-noise PSD of the seven models of seismometers was obtained by using the PDF representation. Based on the mode values of the PDF distribution, the self-noise models of the three channels(UD, EW and NS)of the seven models of seismometers are given respectively.

For the ultra-broadband seismometer CMG-3T-360, in the microseism band(0.1Hz to 1Hz), the self-noise of the horizontal components(EW and NS)is higher than that of the vertical components(UD)and is consistent with the trend of the seismic noise, which may be attributed to misalignment of the horizontal direction between seismometers. In the low frequency band(<0.03Hz), the self-noise level of the horizontal component is higher than that of the vertical component, and small changes in the barometric pressure may lead to higher incoherent noise in the horizontal direction of the sensor at long periods. Compared with the vertical direction, the horizontal direction of the seismometer is more susceptible to air disturbances. At a frequency of 0.005Hz, the instrument self-noise of the horizontal component is close to the seismic background noise, and the instrument self-noise of the horizontal component is the main source of noise recording. Installing a heat and wind shield can effectively reduce the seismometer self-noise in the low frequency band. When using the CMG-3T-360 to observe long-period seismic signals, a shield with both thermal insulation and windproof function is required.

The self-noise level of the short-period seismometer JS-S02 is lower than that of TDV-33S and lower than that of the New Low Noise Model(NLNM)between 0.15Hz and 7Hz. In the UD and EW channels, the self-noise level of TDV-33S is lower than the NLNM model between 0.17Hz and 0.5Hz. The higher instrument self-noise further limits the extraction of long-period seismic signals in the digital recordings of short-period seismometers.

For the broadband seismometer TDV-60B and the very broadband seismometer TDV-120VB, the self-noise levels are basically consistent in the vertical direction and also are higher than those of the broadband seismometer JS-60 and the very broadband seismometer JS-120. In the horizontal direction, the two self-noise levels in the microseism band and the low frequency band show different characteristics, i.e., the self-noise levels of TDV-60B are lower than those of TDV-120VB in the microseism band and higher than those of TDV-120VB in the low frequency band. When the frequency is lower than 0.03Hz, the self-noise levels of TDV-60B and TDV-120VB on the horizontal channels are close to those of JS-60 and JS-120, respectively.

For JS-60 and JS-120, in the vertical direction, the self-noise levels of both are close to CMG-3T-360 in the microseism band. The self-noise level of JS-120 on the vertical channel is lower than 5dB away from CMG-3T-360 in the low frequency band and lies within the 68%confidence interval of the PSD; in the high frequency band(>2Hz), it is higher than CMG-3T-360 confidence interval of the PSD. In the horizontal direction, the self-noise levels of JS-60 and JS-120 are lower than those of CMG-3T-360 between 0.15Hz and 1Hz and in the microseism band, respectively. The self-noise levels of JS-60 on the horizontal channel show good agreement with those of CMG-3T-360 in the high frequency band. The self-noise of JS-120 on NS channel is higher than CMG-3T-360 confidence interval in the low frequency band. When extracting long-period seismic signals, a seismometer with lower noise level in the low frequency band should be selected as much as possible.

Key words: Seismometer, Self-noise, Power Spectral Density, Probability Density Function

王凯明, 余大新, 赵立军, 李文一, 叶庆东. 基于概率密度函数表示的地震计自噪声模型[J]. 地震地质, 2023, 45(5): 1187-1199.

WANG Kai-ming, YU Da-xin, ZHAO Li-jun, LI Wen-yi, YE Qing-dong. SELF-NOISE MODELS OF SEISMOMETERS BASED ON PROBABILITY DENSITY FUNCTION REPRESENTATION[J]. SEISMOLOGY AND GEOLOGY, 2023, 45(5): 1187-1199.

图/表 7

表1 7个型号的地震计噪声功率谱的计算参数

Table 1 Calculation parameters required for PSD of seven models of seismometers

类型	地震计型号	低端截止频率/Hz	电压灵敏度/V·m^-1·s	采样率/sps	数据段长度/h	数据段重叠率/%	数据子段长度/s	数据子段重叠率/%
短周期	JS-S02	0.5	2000	100	1	50	163.84	87.5
短周期	TDV-33S	0.5	2000	100	1	50	163.84	87.5
宽频带	JS-60	0.0167	2000	100	1	50	1310.72	87.5
宽频带	TDV-60B	0.0167	2000	100	1	50	1310.72	87.5
甚宽频带	JS-120	0.00833	2000	100	1	50	1310.72	87.5
甚宽频带	TDV-120VB	0.00833	2000	100	1	50	1310.72	87.5
超宽频带	CMG-3T-360	0.00278	2000	100	4	75	5242.88	87.5

图1 地震背景噪声PSD和CMG-3T-360的自噪声PSD的PDF分布 a、 c和e分别为利用CMG-3T-360 3个通道(UD、 EW和NS)的数据计算的台站背景噪声; b, d和f分别为CMG-3T-360地震计3个通道的自噪声; 红色曲线是经1/8 倍频程平滑的噪声PSD的众值; 绿色曲线为地震计自噪声PDF的第5百分位数; 黑色虚线是EDAS-24GN的噪声PSD曲线

Fig. 1 PDF distributions of seismic ambient noise PSD and CMG-3T-360 self-noise PSD.

图2 背景噪声和CMG-3T-360自噪声功率谱密度的比较

Fig. 2 Comparison of background noise and self-noise by CMG-3T-360 models for UD, EW and NS channel.

图3 Trillium-120PA安装隔热罩和去除隔热罩的自噪声功率谱密度 a 地震计垂直向UD通道的自噪声结果; b 地震计水平向NS通道的自噪声结果; 红色曲线是安装隔热罩计算的自噪声PSD曲线; 蓝色曲线是去除隔热罩后计算的自噪声PSD曲线

Fig. 3 Comparison of Trillium-120PA self-noise models with thermal shield and without thermal shield for UD and NS channel.

图4 7种不同型号地震计(UD通道)自噪声模型的比较黄色区域是CMG-3T-360自噪声功率谱密度68%的置信区间

Fig. 4 Comparison of self-noise models on UD channels for seven different types of seismic instruments.

图5 7种不同型号地震计(EW通道)自噪声模型的比较

Fig. 5 Comparison of self-noise models on EW channels for seven different types of seismic instruments.

图6 7种不同型号地震计(NS通道)自噪声模型的比较

Fig. 6 Comparison of self-noise models on NS channels for seven different types of seismic instruments.

参考文献 23

[1]	葛洪魁, 陈海潮, 欧阳飚, 等. 2013. 流动地震观测背景噪声的台基响应[J]. 地球物理学报, 56(3): 857—868.
	GE Hong-kui, CHEN Hai-chao, OUYANG Biao, et al. 2013. Transportable seismometer response to seismic noise in vault[J]. Chinese Journal of Geophysics, 56(3): 857—868. (in Chinese)
[2]	李小军, 谢剑波, 杨大克, 等. 2019. 基于概率统计的甚宽频带地震计自噪声水平分析[J]. 中国地震, 35(1): 38—52.
	LI Xiao-jun, XIE Jian-bo, YANG Da-ke, et al. 2019. Analysis for self-noise level of the very broadband seismometer using probability statistic[J]. Earthquake Research in China, 35(1): 38—52. (in Chinese)
[3]	许卫卫, 袁松湧, 艾印双, 等. 2017. 多通道相关分析用于宽频带地震仪自噪声检测[J]. 地球物理学报, 60(9): 3466—3474.
	XU Wei-wei, YUAN Song-yong, AI Yin-shuang, et al. 2019. Applying multiple channel correlation technique in the self-noise measurement of broadband seismographs[J]. Chinese Journal of Geophysics, 60(9): 3466—3474. (in Chinese)
[4]	尹昕忠, 陈九辉, 李顺成, 等. 2013. 流动宽频带地震计自噪声测试研究[J]. 地震地质, 35(3): 576—583. doi: 0.3969/j.issn.0253-4967.2013.03.011.
	YIN Xin-zhong, CHEN Jiu-hui, LI Shun-cheng, et al. 2013. The research on self-noise measurement method of moveable broadband seismometer[J]. Seismology and Geology, 35(3): 576—583. (in Chinese)
[5]	薛兵, 朱小毅, 王洪体, 等. 2020. 地震观测仪器进网技术要求: 地震仪(DB/T 22-2020)[M]. 北京: 中国标准出版社: 1—41.
	XUE Bing, ZHU Xiao-yi, WANG Hong-ti, et al. 2020. Technical requirements of the instruments in network for earthquake monitoring: Seismograph(DB/T 22-2020)[M]. Standards Press of China, Beijing: 1—41.
[6]	Ai S, Zheng Y, Xiong C. 2019. Ambient noise tomography across the Taiwan Strait, Taiwan Island, and southwestern Ryukyu Arc: Implications for subsurface slab interactions[J]. Tectonics, 38(2): 579—594. DOI URL
[7]	Berger J, Davis P, Ekström G. 2004. Ambient earth noise: A survey of the global seismographic network[J]. Journal of Geophysical Research: Solid Earth, 109(B11): 1—10.
[8]	Evans J R, Followill F, Hutt C R, et al. 2010. Method for calculating self-noise spectra and operating ranges for seismographic inertial sensors and recorders[J]. Seismological Research Letters, 81(4): 640—646. DOI URL
[9]	Holcomb L G. 1989. A direct method for calculating instrument noise levels in side-by-side seismometer evaluations[R]. US Geological Survey Open File Report, 34: 89—214.
[10]	Holcomb L G. 1990. A numerical study of some potential sources of error in side-by-side seismometer evaluations[R]. US Geological Survey Open File Report, 41: 90—406.
[11]	Inzunza D A, Montalva G A, Leyton F, et al. 2018. Shallow ambient-noise 3D tomography in the Concepción Basin, Chile: Implications for low-frequency ground motions[J]. Bulletin of the Seismological Society of America, 109(1): 75—86. DOI URL
[12]	McNamara D E, Buland R P. 2004. Ambient noise levels in the continental United States[J]. Bulletin of the Seismological Society of America, 94(4): 1517—1527. DOI URL
[13]	Müller T, Zürn W. 1983. Observation of gravity changes during the passage of cold fronts[J]. Journal of Geophysics, 53(1): 155—162.
[14]	Peterson J. 1993. Observations and modeling of seismic background noise[R]. US Geological Survey Open File Report, 95: 93—322.
[15]	Ringler A T, Hutt C R. 2010. Self-noise models of seismic instruments[J]. Seismological Research Letters, 81(6): 972—983. DOI URL
[16]	Ringler A T, Hutt C R, Evans J R, et al. 2011. A comparison of seismic instrument noise coherence analysis techniques[J]. Bulletin of the Seismological Society of America, 101(2): 558—567. DOI URL
[17]	Ringler A T, Sleeman R, Hutt C R, et al. 2014. Seismometer Self-Noise and Measuring Methods [G]// Beer M, Kougioumtzoglou I, Patelli E, et al.(eds). Encyclopedia of Earthquake Engineering. Berlin, Heidelberg: Springer Berlin Heidelberg: 1—13.
[18]	Shapiro N M, Campillo M, Stehly L, et al. 2005. High resolution surface wave tomography from ambient seismic noise[J]. Science, 307(5715): 1615—1618. DOI URL
[19]	Sleeman R, Melichar P. 2012. A PDF representation of the STS-2 self-noise obtained from one year of data recorded in the Conrad Observatory, Austria[J]. Bulletin of the Seismological Society of America, 102(2): 587—597. DOI URL
[20]	Sleeman R, van Wettum A, Trampert J. 2006. Three-channel correlation analysis: A new technique to measure instrumental noise of digitizers and seismic sensors[J]. Bulletin of the Seismological Society of America, 96(1): 258—271. DOI URL
[21]	Tasič I, Runovc F. 2012. Seismometer self-noise estimation using a single reference instrument[J]. Journal of Seismology, 16(2): 183—194. DOI URL
[22]	Welch P D. 1967. The use of fast Fourier transform for the estimation of power spectra: A method based on time averaging over short, modified periodograms[J]. IEEE Transactions on Audio and Electroacoustics, 15(2): 70—73. DOI URL
[23]	Zürn W, Widmer R. 1995. On noise reduction in vertical seismic records below 2mHz using local barometric pressure[J]. Geophysical Research Letters, 22(24): 3537—3450. DOI URL