SUMMARY Seismic attenuation measurements, especially those obtained from coda decay analysis, are becoming a key data source for the characterization of the heterogeneous Earth due to their sensitivity to small-scale heterogeneities. However, the relation between the scattering attenuation measured from coda waves and physical rock properties is still unclear. The goal of this study is to identify the main petrophysical and mineralogical factors controlling coda attenuation in volcanic rocks at the laboratory scale, as a necessary step before modelling seismic waves in real volcanic media. Coda wave attenuation was estimated from ultrasonic S-wave waveforms. To quantify the heterogeneity of the rocks and link them with this attenuation parameter, we performed several categorizations of the pore and grain systems of volcanic samples. Considering that seismic attenuation in rock samples can be modelled using the framework of wave propagation in random media, a statistical analysis of shear wave velocity fluctuations was performed: this analysis gives correlation lengths ranging from 0.09 to 1.20 mm, which represents the length scale of heterogeneity in the samples. The individual evaluation of the pore space and mineral content revealed that the pores of the samples (characterized by large vesicles) have a bigger effect than the grains on the heterogeneity level. We have developed a framework where intrinsic properties of the host rocks drive seismic attenuation by correlating the petro-mineralogical characteristics obtained from image data processing and analysis, with the coda attenuation measured at ultrasonic frequencies. There is conclusive evidence that porosity alone is not the primary controller of coda attenuation: it is also changed by the alteration level (i.e. oxidation, coating of the vesicles, secondary minerals) and the size of grains and pores. Among all the parameters analysed, it appears that the pore space topology is the main contributor to scattering attenuation in the volcanic samples.

中文翻译：

---

火山岩样品尾声衰减的石油矿物学控制

总结 地震衰减测量，特别是从尾波衰减分析中获得的那些，由于它们对小尺度异质性的敏感性，正在成为表征异质地球的关键数据源。然而，尾波测量的散射衰减与岩石物理性质之间的关系仍不清楚。本研究的目的是确定在实验室尺度上控制火山岩尾波衰减的主要岩石物理和矿物学因素，这是在真实火山介质中模拟地震波之前的必要步骤。从超声波 S 波波形估计尾波衰减。为了量化岩石的非均质性并将它们与衰减参数联系起来，我们对火山样品的孔隙和颗粒系统进行了几种分类。考虑到岩石样本中的地震衰减可以使用波在随机介质中传播的框架进行建模，对剪切波速度波动进行了统计分析：该分析给出的相关长度范围为 0.09 到 1.20 mm，代表了非均质的长度尺度在样本中。对孔隙空间和矿物含量的单独评估表明，样品的孔隙（以大囊泡为特征）对非均质性水平的影响大于晶粒。我们开发了一个框架，通过将图像数据处理和分析获得的岩石矿物学特征与在超声波频率下测量的尾波衰减相关联，主岩的内在特性驱动地震衰减。有确凿的证据表明，孔隙度本身并不是尾声衰减的主要控制因素：它还受到蚀变水平（即氧化、囊泡涂层、次生矿物）以及颗粒和孔隙大小的影响。在所有分析的参数中，孔隙空间拓扑似乎是火山样品中散射衰减的主要贡献者。