Abstract:

In recent years, both governmental agencies and the public have placed increasing emphasis on the timeliness of post-earthquake trend analysis. However, during the early stages following an earthquake—particularly within the first few hours(the so-called “zero-hour” period)—the severe lack of observational data necessitates reliance on prior statistical information, such as the historical proportions of different earthquake sequence types. Accordingly, systematic statistical analyses of the proportional distribution of seismic sequence types and the characteristics of the largest aftershocks in different regions, based on historical earthquake sequence data, can provide essential reference data for the rapid identification of early sequence types and post-earthquake trends during this critical prediction bottleneck. Owing to the combined effects of regional tectonic settings, subsurface media, and stress fields, seismic activity in Yunnan is highly complex, resulting in diverse and complicated earthquake sequence types.

Using data from the China Earthquake Cases and monthly catalogs of the Yunnan Seismic Network, this study compiled 152 earthquake sequences with M≥5.0 that occurred between 1966 and 2023 in Yunnan and its adjacent areas(20°~29.5°N, 97°~106°E). Earthquake sequences were classified according to the magnitude difference ΔM=M₀-M₁ between the largest and second-largest events in each sequence into three types: multiple mainshock type(MMT, ΔM<0.6), mainshock-aftershock type(MAT, 0.6≤ΔM≤2.4), and isolated earthquake type(IET, ΔM>2.4). The influence of mainshock magnitude and rupture style on sequence type was analyzed. For forecasting the largest aftershock—a key issue in post-earthquake trend analysis—statistical investigations were conducted from three perspectives: the relationship between mainshock magnitude and the largest aftershock magnitude, the time interval between the mainshock and the largest aftershock, and the spatial distribution characteristics of magnitude differences. In addition, the spatial distribution patterns of earthquake sequence types in Yunnan and surrounding regions were examined, with particular focus on the distribution of MMT sequences and their relationship with regional tectonic structures.

The results indicate that: 1)Earthquake sequences in the study area are dominated by the MAT type, followed by MMT, with IET being the least common. Within the same sequence type, the proportions of MAT and MMT increase with increasing mainshock magnitude, whereas the proportion of IET decreases. Among different rupture styles, IET sequences are absent in normal-fault earthquakes, while thrust-fault earthquakes exhibit relatively high proportions of MMT, mainly distributed in northeastern Yunnan. 2)Linear regression analysis between mainshock magnitude and the magnitude of the largest aftershock shows the strongest correlation for MMT sequences, followed by MAT sequences, whereas IET sequences display the greatest scatter. The occurrence time of the largest aftershock is related to both sequence type and mainshock magnitude. The spatial distribution of D₁ values shows significant regional differences, with the Honghe and Xiaojiang faults acting as boundaries: D₁ values are highest east of the Xiaojiang Fault and the Zhaotong-Ludian Fault, and lowest west of the Red River Fault. 3)The spatial distribution of earthquake sequences in Yunnan and its adjacent regions exhibits statistically significant regional characteristics. Overall, MMT sequences are most prevalent in western Yunnan, followed by northeastern Yunnan, with relatively high proportions also observed in northwestern and southwestern Yunnan. Central Yunnan displays the simplest composition of sequence types, whereas western Yunnan shows the greatest complexity.

The distribution characteristics of MMT sequences can be broadly summarized as follows: 1)MMT commonly occurs in source environments characterized by complex seismogenic fault systems, such as conjugate or multiple intersecting fault sets, the interweaving of concealed structures, or the coupling of fault systems at different depths; 2)newly formed fault zones, exemplified by the Longling-Lancang fault zone, which developed after cutting through folded geological bodies of older faults and consists of numerous discontinuous, small-scale, parallel, clustered, or obliquely oriented secondary faults, are prone to generating MMT sequences with relatively large magnitudes; 3)upper-crustal low-velocity zones, as the spatial distribution of sequence types is closely related to the deep structural environment, with MAT sequences mainly occurring in high-velocity zones or transitional zones between high- and low-velocity regions, whereas MMT sequences are more frequently associated with low-velocity zones in the upper crust; 4)areas near the epicenters of historical large earthquakes, where in central and northwestern Yunnan MMT sequences tend to cluster around the epicentral regions of past major events; and 5) in recent years, earthquake sequences induced by human industrial activities, such as reservoir-induced seismicity and hydraulic fracturing, have exhibited increasingly complex sequence types. It should be noted that the first four characteristics are not mutually exclusive in the seismogenic processes of MMT sequences; in many cases, the source environments of MMT events may simultaneously involve multiple factors, particularly the first three tectonic features.

Key words: earthquake sequence type, maximum aftershock magnitude, multiple mainshock type, multi sequence spatial distribution

摘要：

本文根据云南及邻区1966年以来152组M≥5.0地震序列统计结果, 对地震序列类型及其空间分布特征、 最大余震震级和时空分布特征进行了分析, 结果表明: 1)研究区域地震序列以主余型为主, 多震型次之, 孤立型最少; 同一序列类型中, 随着地震震级增大, 主余型、 多震型地震所占的比例增加, 孤立型逐渐减少; 对于不同破裂类型, 正断型地震中, 没有孤立型地震序列; 逆冲型地震中, 多震型的比例较高, 且主要分布在滇东北地区。2)主震与最大余震震级线性回归结果显示, 多震型线性相关性最好, 其次是主余型, 孤立型序列数据离散程度较高; 最大余震的发生时间与序列类型和主震震级有一定关系; D₁值以红河断裂和小江断裂为界, 具有明显的空间分布差异特征。3)地震序列空间分布具有较明显区域性特征, 多震型比例最高的是小滇西, 其次是滇东北; 地震序列类型最单一的区域是滇中, 相对复杂的区域为小滇西。4)多震型地震的孕震环境可能与复杂共轭的构造环境、 新生断裂、 上地壳低速区及历史大地震分布有一定的关系。

关键词: 地震序列类型, 最大余震震级, 多震型地震序列, 序列空间分布特征