搜集2014-2015年曹妃甸测震台网记录的各类爆破与天然地震事件，基于Matlab软件，编制程序实现短时傅里叶变换，对比分析核爆炸、化学爆炸、矿爆炸与天然地震的特征，分析频率随时间的变化规律。研究结果表明，天然地震与爆破在频率范围、频率分布和时频聚集性等方面具有各自特点，主要表现在：天然地震的时频图"多峰"且复杂，频率分布范围更广，频率成分集中在S波段；人工爆破的时频图"少峰"且简单，频率范围较窄，频率成分集中在P波段。本文的研究成果能够为今后区分天然地震和爆破提供依据。
Based on Matlab software, we developed a program to achieve the short time Fourier transforms. We collected all kinds of seismic blasting and natural seismic events that recorded by Caofeidian Seismic Network from 2014 to 2015. Using time frequency analysis method, the nuclear explosions, mine explosions, chemical explosions and natural earthquakes that were recorded by Caofeidian Seismic Network are studied. The instantaneous frequency is calculate. Research results show that natural earthquakes and blasting have their own characteristics in frequency range, frequency distribution and time frequency aggregation. Time frequency map of natural earthquake is multimodal, the frequency distribution is wider, and frequency components gather in S band. The time-frequency map of the artificial blasting is less peak and simple, the frequency range is narrow, and the frequency components gather in P band. This study provides a basis for distinguishing natural earthquakes and explosions in the future.
2017,38(5): 39-43 收稿日期：2016-09-13
DOI：10.3969/j.issn.1003-3246.2017.05.007
基金项目：河北省地震局青年激励项目"曹妃甸测震台网天然地震与人工爆破的时频分析"；河北省地震局科技星火计划项目"滦县井水位数字化观测及监测效能评估"（项目编号：DZ20170510055）
作者简介：荣伟健(1991-),男,助理工程师,2015年毕业于防灾科技学院,从事地震监测工作
参考文献：
安镇文,魏富胜. 震源性质的时频分析与事件识别[J]. 地震地磁观测与研究,2011,32(1):1-9.
崔鑫,许力生,许忠淮,等. 小地震与人工爆破记录的时频分析[J]. 地震工程学报,2016,38(1):71-78.
韩绍卿,李夕海,安跃文,等. 核爆、化爆、地震识别研究综述[J]. 地球物理学进展,2010,25(4):1 206-1 218.
刘芳彤,杜春鸿. 朝鲜地区地震与爆破波形识别研究[J]. 防灾减灾学报,2015,31(2):58-66.
王新刚,郭琦南. 时频分析方法在探测信号处理中的应用[J]. 电脑知识与技术,2013,9(26):5 973-5 975.
万永革. 数字信号处理的Matlab实现[M]. 北京:科学出版社,2007:406-411.
张帆,朱新运,熊丹,等. 基于非线性时频分析的地震和爆破识别[J]. 华南地震,2014,34(2):56-63.
赵建明,任佳,符泽宇,等. 曹妃甸地震台电磁扰动异常分析[J]. 地震地磁观测与研究,2015,36(4):26-29.
