Investigating Spectral Distortion of Local Volcano Infrasound by Nonlinear Propagation at Sakurajima Volcano, Japan
Investigating Spectral Distortion of Local Volcano Infrasound by Nonlinear Propagation at Sakurajima Volcano, Japan
复制标题
DOI：
10.1029/2019jb018284
复制
发表时间：
2020-03
期刊：
影响因子：
--
通讯作者：
S. Maher;R. Matoza;C. Groot‐Hedlin;K. Gee;D. Fee;A. Yokoo
中科院分区：
文献类型：
--
作者：
S. Maher;R. Matoza;C. Groot‐Hedlin;K. Gee;D. Fee;A. Yokoo
火山喷发产生的声波可用于推断重要的震源动力学信息，然而，即使在局部记录距离（<15千米）内，声震源时间函数在传播过程中也可能发生扭曲。通过改进对传播效应的约束，可以降低震源估计中的不确定性。我们旨在量化非线性传播过程中波陡化所导致的潜在扭曲，以提高火山声震源预测的准确性。我们假设波陡化会导致频谱能量从主震源频率转移。为了验证这一点，我们将先前开发的基于单点、频域、四谱密度的非线性指标应用于日本樱岛火山武尔卡诺式喷发事件的30个声学信号，这些信号来自2013年由5个次声站在2.3 - 6.2千米范围内收集的8天数据集。我们使用包含刚性地形、风和非线性传播的二维轴对称有限差分方法对这些结果进行建模。平坦地面的模拟结果表明，对于樱岛火山的振幅，波陡化会导致高达约2分贝（震源水平的1%）的累积向上频谱能量转移。因此，对非线性传播进行校正可能会为震源参数估计的准确性提供有价值的二阶改进。然而，包含风和地形的模拟在指标频谱中引入了几分贝数量级的变化。传播过程中或在震源处产生的非随机相位关系可能会被误解为非线性频谱能量转移。因此，该非线性指标最适用于较小的震源 - 接收器距离（例如，<2千米）以及具有简单震源（例如，富含气体的斯特龙博利式喷发）和地形的火山。
Sound waves generated by erupting volcanoes can be used to infer important source dynamics, yet acoustic source‐time functions may be distorted during propagation, even at local recording distances ( < 15 km). The resulting uncertainty in source estimates can be reduced by improving constraints on propagation effects. We aim to quantify potential distortions caused by wave steepening during nonlinear propagation, with the aim of improving the accuracy of volcano‐acoustic source predictions. We hypothesize that wave steepening causes spectral energy transfer away from the dominant source frequency. To test this, we apply a previously developed single‐point, frequency domain, quadspectral density‐based nonlinearity indicator to 30 acoustic signals from Vulcanian explosion events at Sakurajima Volcano, Japan, in an 8‐day data set collected by five infrasound stations in 2013 with 2.3‐ to 6.2‐km range. We model these results with a 2‐D axisymmetric finite‐difference method that includes rigid topography, wind, and nonlinear propagation. Simulation results with flat ground indicate that wave steepening causes up to ∼ 2 dB (1% of source level) of cumulative upward spectral energy transfer for Sakurajima amplitudes. Correction for nonlinear propagation may therefore provide a valuable second‐order improvement in accuracy for source parameter estimates. However, simulations with wind and topography introduce variations in the indicator spectra on order of a few decibels. Nonrandom phase relationships generated during propagation or at the source may be misinterpreted as nonlinear spectral energy transfer. The nonlinearity indicator is therefore best suited to small source‐receiver distances (e.g., < 2 km) and volcanoes with simple sources (e.g., gas‐rich strombolian explosions) and topography.