论文部分内容阅读
用多元成分数据统计分析中的端元估计算法，处理混合源沉积物地球化学数据，定量判识沉积物物源端元的化学成分和比例。将该方法应用于已知端元的人工数据，端元的化学成分和含量的估计值与已知值非常接近。实验结果说明，这种成分数据分析算法是定量判识沉积物物源端元的有效方法。用此方法处理冲绳海槽中段表层沉积物地球化学数据集，成功地定量估计出陆源物质、火山源物质和生物源物质的平均化学成分；同时定量计算出该区底质沉积物中陆源物质、火山源物质和生物源物质的百分含量。最后，将实算结果与矿物学判识的物源分布进行对比，二者总趋势一致，前者的局部变化大。此方法用于物源组成的定量估计工作，既经济又快速。
Using the endmember estimation algorithm in the statistical analysis of the multivariate composition data, the geochemical data of the mixed source sediment are processed to quantitatively identify the chemical composition and proportion of the source end of the sedimentary material. Applying this method to the artificial data of the known endmember, the chemical composition and content of the endmember are estimated very close to the known values. The experimental results show that this method of component data analysis is an effective method to quantitatively identify the source end of sediments. By using this method, the geochemical data sets of the surface sediments in the middle Okinawa Trough are processed and the average chemical compositions of the terrigenous, volcanic and biogenic materials are quantitatively and quantitatively estimated. At the same time, the terrigenous sediments, Percentage of volcanic and biogenic material. Finally, comparing the actual results with the provenance distribution of mineralogical analysis, the general trend of the two is the same, and the local variation of the former is large. This method is used for the quantitative estimation of the composition of provenances, which is both economical and rapid.