Abstract—The digital seismic network in Kamchatka deployed in 2006–2010 provided a fundamental possibility for calculating seismic moment tensor (SMT) of Kamchatka earthquakes from broadband waveforms recorded by regional stations. The paper describes the method for calculating SMT for the earthquakes with pure double couple (DC) mechanism based on the DC tensor decomposition and explicit partial linear inversion in two (of four) variables. This allows the least square objective function to be expressed in terms of two angles specifying the orientation of the neutral (zero) axis of the DC tensor. The inversion with respect to the angles is conducted using the Levenberg–Marquardt iterative method. This technique was used for calculating focal mechanisms, moment magnitudes, and equivalent point-source depths for 31 Kamchatka earthquakes with moment magnitudes M_w = 4.3–6.2. The comparison of the obtained estimates with the corresponding GCMT catalog data which exist for 19 events has shown fairly reasonable agreement between the results. The angle K characterizing the discrepancy between the DC mechanism calculated directly from the waveforms in our study and the best DC mechanism constructed from GCMT null-trace tensor is at most 25° for 16 events of 19. The moment magnitudes are underestimated by ~0.1 relative to the GCMT counterparts and the depths are underestimated by 8 km on average.

摘要— 2006-2010 年部署的堪察加数字地震台网为根据区域台站记录的宽带波形计算堪察加地震的地震矩张量 (SMT) 提供了基本可能性。本文描述了基于 DC 张量分解和显式部分线性反演在两个（四个）变量中计算纯双偶 (DC) 机制地震 SMT 的方法。这允许根据指定 DC 张量的中性（零）轴方向的两个角度来表示最小二乘目标函数。使用 Levenberg-Marquardt 迭代方法进行角度的反演。该技术用于计算焦点机制、力矩大小、M _w = 4.3–6.2。将获得的估计值与存在于 19 个事件的相应 GCMT 目录数据进行比较，结果之间显示出相当合理的一致性。表征直接从我们研究中的波形计算的 DC 机制与由 GCMT 零迹张量构建的最佳DC 机制之间差异的角度K对于 19 个事件中的 16 个事件最多为 25°。力矩幅度被相对低估了约 0.1与 GCMT 同行相比，深度平均被低估了 8 公里。