年轻火山岩激光40Ar/39Ar定年中的n值与MSWD值

MSWD and n Value in Laser Ablation ^(40)Ar/^(39)Ar Isochron Dating for Young Volcanic Rocks

激光熔蚀40Ar/39Ar等时线法是第四纪年轻火山岩样品精细定年的新方法,该方法处于逐步完善过程中。针对怎样判定获得的年龄结果是否可靠的问题,通过对数据处理过程中所用参数的分析,认为样品属性是决定年轻火山岩精细定年结果质量和可靠性的关键因素,与样品测试数n值相关的加权均方差MSWD(meansquareweighteddeviation)是判断同一样品各测试点的数据是否符合正态分布,能否用来客观地评价测试结果,进而评估样品是否适合进行相关测试,结果是否可靠的重要指标。其次,属于同一真实年龄值样品的MSWD在99.7%置信区间的取值范围可通过计算加以确认,并提出对于表面年龄和等时线年龄数据,必须定量地考察相应的MSWD值,判定在具有足够样品测试数(足够大的n值)的情况下,是否符合概率统计的正态分布,确定样品是否对应同一年龄真实值。在MSWD值达不到理想值1.0的情况下,可以利用表面年龄概率密度曲线,区分测试样品内部包含的非同源或非等时或封闭不好体系的数据,再结合MSWD与n值关系,适量地去除部分偏离正态分布的测试数据,识别出可能归属多个年龄真实值的测试数据组,并以此计算不同数据组的反等时线年龄,获得年轻火山岩喷发的准确年龄。

The laser ablation 40Ar/39Ar method makes it possible to date quanterary volcanic rocks precisely.So far,it lacks an assessment criterion that can be used to estimate the reliability of test data and geochronological information.The mean square weighted deviation(MSWD)and the number of measurement(n value)are used to be the crucial parameter to evaluate sample properties and experimental standard.When applying laser ablation 40Ar/39Ar method to date young volcanic rocks,a best-fit 40Ar/39Ar isochron diagram requires isotopic homogeneity of the sample and permanent closure of the isotope system at the same time.While MSWD,which is related to n value,is a parameter to determine whether the data are located in the same normal distribution,or in other words,whether the sample has a unified true age,it is applicable to use MSWD and n value to estimate the level of confidence placed on the computed mean.The critical values of MSWD for the 99.7%confidence interval using differing degrees of freedom are calculated in this paper.It is important to note that MSWD is available in evaluating dating result of both apparent and isochron ages.For the samples which tend to have heterogeneous initial 40Ar/36Ar ratios or different closure times,MSWD combined with the probability density of apparent ages can be used to distinguish different groups of samples,calculate the isochron ages respectively,and interpret the result after accounting for the various parameters to evaluate the data quality.