Use of Fractal Analysis for the Discrimination of Volcanic Particles

年轻的火山颗粒是现代和古代火山地体中火山碎屑岩的重要组成部分。火山碎屑沉积的识别通常依赖于对这种颗粒的分析。尽管在将颗粒形态的定性方面与不同类型的源喷发和碎裂过程联系起来方面取得了重大进展，但颗粒分类的定量方法仍然难以找到。我们建议使用分形几何学作为火山颗粒判别的一种定量方法。火山颗粒由于其复杂的形状和边界上存在重入物，特别适合于应用分形分析。我们开发了一种基于火山颗粒轮廓中离散分形元素的量化和统计分析的新技术。这项技术的初步工作已显示出对不同火山颗粒进行定量区分的极好潜力。我们建议对一套特征良好的样品进行分形分析，以建立一个代表各种火山碎屑沉积的分形特征矩阵。这些样品将包括从玄武岩到流纹岩的岩浆类型和破碎机制，包括原生挥发性脱气作用、镜质岩浆作用和海水淬火造成的颗粒化。此外，样品将从不同的运移机制产生的沉积物中进行分析，例如火山碎屑坠落、流动和涌动。我们预计，这项研究的结果将在火山颗粒形态的量化方面取得重大进展，并导致统计上区分不同来源的颗粒的手段。这项技术对于寻求从所形成的沉积物特征来解释古代火山活动性质的工作人员来说将是很有价值的。此外，对颗粒形态的详细定量分析可能会导致对复杂爆炸喷发过程中各种破碎过程的作用有新的见解，包括初级挥发分的脱气和与外部水的相互作用。

CareyEAR-0086963Juvenile volcanic particles form an important component of pyroclastic rocks in recent and ancient volcanic terrains. The identification of volcaniclastic deposits often relies on an analysis of such particles. Although significant progress has been made in linking qualitative aspects of particle morphology to different types of source eruptions and fragmentation processes, a quantitative method of particle classification has remained elusive. We propose to investigate the use of fractal geometry as a quantitative method of volcanic particle discrimination. Volcanic particles are particularly well suited for the application of fractal analysis owing to their complex shapes and presence of re-entrants on their boundaries. We have developed a new technique based on the quantification and statistical analysis of discrete fractal elements in the outlines of volcanic particles. Preliminary work on this technique has shown excellent potential for quantitative discrimination of different volcanic particles. We propose to carry out fractal analysis on a suite of well-characterized samples to build up a matrix of fractal characteristics that will be representative of a wide variety of volcaniclastic deposits. The samples will encompass magma types from basalt to rhyolite and fragmentation mechanisms that include primary volatile degassing, phreatomagmatism, and granulation by quenching with seawater. In addition, samples will be analyzed from deposits produced by different transport mechanisms such as pyroclastic fall, flow and surge. We anticipate that the results of the study will provide significant advances in the quantification of volcanic particle morphology and lead to the means of statistically discriminating particles of different origin. Such a technique will be valuable to workers seeking to interpret the nature of ancient volcanic activity from the characteristics of their resulting deposits. Furthermore, the detailed quantitative analysis of particle morphology may lead to new insights about the roles of various fragmentation processes during complex explosive eruptions involving both the degassing of primary volatiles and the interaction with external water.