海洋沉积物粒度参数3种计算方法的对比研究

A comparative study on the grain-size parameters of marine sediments derived from three different computing methods

海洋沉积物粒度参数是反映沉积物物质来源及水动力条件的敏感性指标。目前国内对海洋沉积物粒度参数计算方法并不完全统一,制约了粒度数据的整合和使用。利用长江口外近海348个表层沉积物粒度数据,分别使用Folk和Ward的图解法公式、Collias等的矩法公式和McManus的矩法公式进行了粒度参数的计算。运用相关分析和基本统计分析等方法,讨论了不同方法计算结果间的差异和影响因素。结果显示:图解法与矩法计算的平均粒径和分选系数分别呈显著线性相关,且不受沉积物物质来源的影响。对于偏态和峰态,3种不同方法计算结果的相关性较低,甚至表现为不相关。不同方法计算的沉积物粒度参数的差异性均与沉积物类型和频率曲线分布形态有关。当物质组成均匀、频率曲线近正态分布时,不同方法计算的粒度参数的相关性显著提高。当沉积物较粗或较细时,粒度分布偏离正态且具有明显尾部特征,用不同方法计算结果的差异会随之增大。对3种不同公式综合对比认为Collias等的矩法公式反映沉积物粒度分布特征更加灵敏、可靠,建议在进行资料整合时采用。

Grain-size parameters of marine sediments are the sensitive indexes for reflecting the sediment sources and hydrodynamic conditions.At present,the various methods for calculating the grain-size parameters of marine sediments partially restrict the integration and availability of grain-size data.Three hundred and four-eight surface sediments grain-size data from the offshore of the Changjiang River Estuary are used to calculate the grain-size parameters,respectively adopting the graphical method of Folk and Ward,the moment method of Collias et al.and the moment method of McManus.Correlation analysis and statistical analysis methods are employed to detect the differences of results from the above three different methods and their influencing factors.The mean grain-size and sorting coefficient derived from the graphical method and the moment methods are significantly correlative and stable,which are not affected by sediment sources.On the contrary,the values of skewness and kurtosis calculated from three different methods indicate low correlation or even no relativity.The results show that the differences of grain-size parameters calculated from the three methods depend both on sediment types and their form of distribution curve.When sediments components are uniform with a normal frequency curves,there are high correlation coefficients amount the grain-size parameters from different methods,While for the case of the coarse or fine sediments with the abnormal frequency curves of a long tail feature,the different of grain-size parameters using difference methods are increased.The results suggest that Collias et al.moment formulas are recommended to be used in the process of grain-size integration for their sensitivity and reliability in reflecting the grain-size distribution characteristic.