Infrasound records at Grey glacier, Patagonia, Chile.

A four-element infrasound array was deployed at Grey Glacier (Torres del Paine, Patagonia, Chile) between February 4–11, 2025. This dataset includes infrasound array data recorded by the four sensors in MiniSEED (.mseed) format, along with detection results from array processing provided in MATLAB (.mat) format.The GREY array is composed of four Item differential pressure transducers with a sensitivity of 400 mV/Pa, operating in the frequency band of 0.01–200 Hz and a full-scale pressure range of ±25 Pa. Each sensor is connected to a 24-bit Güralp Minimus digitizer, sampling at 100 Hz and synchronized with GPS time.Here the coordinates of the four sensors and the corresponding file name:Channels Latitude (S) Longitude (W) Altitude (m) FilenameCH1 50,986 73,27372 80 FI.GREY.01.HDFCH2 50,98554 73,27427 80 FI.GREY.02.HDFCH3 50,98583 73,27364 68 FI.GREY.03.HDFCH4 50,98573 73,27385 65 FI.GREY.04.HDF<br>Array processing was performed using multi-channel cross-correlation techniques in the time domain, applied to moving time windows across all array elements, with the same method of Ulivieri et al., 2011. The data were bandpass-filtered between 2 and 20 Hz and analyzed using 5-second moving windows with 90% overlap (shift = 0.5 s). The detections characterize the infrasound wavefield in terms of pressure amplitude, back-azimuth and apparent velocity. The file <i>GREY_detection.mat </i>contains the detections -the result of the array processing. This files contains two structures: <i>data</i> and <i>info</i>, which store processed infrasound detection results and associated processing parameters.Structure: <i>data</i>txx: Detection time in MATLAB datenum format (UTC).prs: Peak acoustic pressure values in the time window of analysis (Pascal).cc: Coherence values, indicating the degree of cross-correlation analysis (0-1).rs: Time Residual values from delays times (samples).Fp: Peak frequency values in the window of analysis (Hz).azz: Back-azimuth (° from North) indicating source direction.slw: Apparent velocity (m/s), related to wave propagation speed on the plane described by the array.Structure:<i> info</i>maxl: Maximum time lag in cross-correlation analysis (samples).freq: Frequency band (Hz) used for filtering the signals.win: Time window length for analysis (seconds).shift: Time step of moving time windows (seconds).minR: Minimum allowed time residual (seconds)sensors: Active sensors used for the array processing (1- active; 0-non active).The MATLAB script <i>detection_per_hour_polar_plot.m</i> visualizes infrasound detection data from the <i>Grey_Detections.mat</i> file, including basic filtering and plotting. It performs the following steps:Data Cleaning and Filtering: Removes spurious detections based on residual (&lt;20), slowness (200–500 s/km), and coherence (&gt;0.1).Applies two levels of coherence filtering: initial and stricter (&gt;0.7) for comparison.Time Series Plots: Plots pressure and back-azimuth over time, before and after filtering. Uses color and marker size to distinguish raw and filtered data.Hourly Detection Histogram: Computes and displays hourly counts of detections on a logarithmic scale.Back-Azimuth Distribution: Creates polar histograms showing the directional distribution of detections, both unfiltered and filtered by high coherence.<br>2025年2月4日至11日期间，一套四元次声阵列（infrasound array）部署于智利巴塔哥尼亚托雷斯德尔佩恩的灰冰川（Grey Glacier）。 本数据集包含由四台传感器采集的次声阵列数据，格式为MiniSEED（.mseed），同时附带阵列处理得到的检测结果，格式为MATLAB（.mat）。 该GREY阵列由四台灵敏度为400 mV/Pa的差分压力传感器（differential pressure transducers）组成，工作频段为0.01–200 Hz，满量程压力范围为±25 Pa。每台传感器均连接至24位Güralp Minimus数字化仪，采样率为100 Hz，并通过GPS时间实现同步。 以下为四台传感器的坐标及对应文件名： | 通道 | 南纬纬度 | 西经经度 | 海拔（米） | 文件名 | |------|----------|----------|------------|--------| | CH1 | 50.986 | 73.27372 | 80 | FI.GREY.01.HDF | | CH2 | 50.98554 | 73.27427 | 80 | FI.GREY.02.HDF | | CH3 | 50.98583 | 73.27364 | 68 | FI.GREY.03.HDF | | CH4 | 50.98573 | 73.27385 | 65 | FI.GREY.04.HDF | 阵列处理采用时域多通道互相关技术，对所有阵列单元的移动时间窗进行处理，方法与Ulivieri等人2011年的研究一致。数据经2–20 Hz带通滤波后，采用时长5秒、重叠率90%（步长=0.5秒）的移动窗进行分析。检测结果以压力振幅、反方位角（back-azimuth）和视速度（apparent velocity）表征次声波场。 文件`GREY_detection.mat`包含阵列处理得到的检测结果，该文件包含两个结构体：`data`与`info`，分别存储处理后的次声检测结果及相关处理参数。 ### 结构体`data` - `txx`：采用MATLAB日期序列号（MATLAB datenum format）格式表示的检测时间（UTC）。 - `prs`：分析时间窗内的峰值声压值（单位：帕斯卡）。 - `cc`：相干性（coherence）数值，表征互相关分析的相关程度（取值范围0–1）。 - `rs`：时延时间残差（单位：采样点）。 - `Fp`：分析时间窗内的峰值频率（单位：Hz）。 - `azz`：反方位角（单位：°，以正北为基准），指示信号源方向。 - `slw`：视速度（单位：m/s），表征次声波在阵列所在平面的传播速度。 ### 结构体`info` - `maxl`：互相关分析中允许的最大时延（单位：采样点）。 - `freq`：信号滤波所用的频段（单位：Hz）。 - `win`：分析所用时间窗的时长（单位：秒）。 - `shift`：移动时间窗的步长（单位：秒）。 - `minR`：允许的最小时间残差（单位：秒）。 - `sensors`：阵列处理中使用的有效传感器标识（1表示有效，0表示无效）。 MATLAB脚本`detection_per_hour_polar_plot.m`可可视化`GREY_Detections.mat`文件中的次声检测数据，包含基础滤波与绘图功能，具体执行步骤如下： 1. **数据清洗与滤波**：基于残差（<20）、慢度（slowness，200–500 s/km）及相干性（>0.1）剔除异常检测结果；同时应用两级相干性滤波（初始阈值与严格阈值>0.7）用于对比分析。 2. **时间序列绘图**：绘制滤波前后的声压与反方位角随时间变化曲线，通过颜色与标记尺寸区分原始数据与滤波后数据。 3. **逐小时检测直方图**：以对数尺度计算并展示逐小时的检测次数分布。 4. **反方位角分布**：生成极坐标直方图，展示未滤波及经高相干性筛选后的检测结果的方位分布。