• 中文核心期刊
  • 中国科技核心期刊
  • CSCD来源期刊
  • DOAJ 收录
  • Scopus 收录

地球化学普查样品实验室分析异常点抽检改进方法

Improvement of the Anomaly Spot-checking Method in Sample Analysis during a Geochemical Reconnaissance Survey

  • 摘要: 在地球化学样品多元素分析过程中,为了实现地球化学异常点抽查和整批测试质量监控的双重目的,本文提出了一种基于地球化学异常统计的多区间分元素组异常点抽样方法。该方法主要包括:①按照正态分布或对数正态分布假设,对各元素,将该元素的测量值集,划分为离群低值区间、累积频率分布区间(有多个区间)和离群高值区间,统计每个样品中各元素测量值所属区间;②将测量元素分为几个元素组,统计各元素组中每个样品的总异常元素数、高异常元素数和低异常元素数;③对Au元素组,设置各个区间的抽检比例,随机抽取异常点抽检样品;对其他元素组,按照“总异常元素数”为第一顺序、“高异常元素数”为第二顺序和“低异常元素数”为第三顺序,按降序排序所有样品,顺序抽取异常点抽检样品;④必要时进行分析批补充抽检样品。该方法兼顾了地球化学异常和实验室测试“异常”的抽查;同时考虑非异常点抽查和分析批平衡抽检,有效解决了有限抽样数量和抽样代表性不足两者之间的矛盾;并开发了EXCEL应用程序,应用于实验室地球化学样品分析异常点抽检取得了满意效果。

     

    Abstract: In order to meet the double goals of picking up real geochemical anomalies and monitoring inner laboratory analytical quality during multiple element analysis of geochemical samples, a multi-interval and element-grouping sampling procedure (MIEGSP) based on the statistical method of geochemical anomalies is proposed. The MIEGSP mainly includes:(1) the set of each element's (logarithmic) measurements supposed to obey normal distribution and is divided into successive intervals including low/high value outlier ones and cumulative frequency distributions. Each measurement is assigned to a specific interval. (2) Target elements are grouped, total anomaly number (TAN), high anomaly number (HAN) and low anomaly number (LAN) in each group are calculated. (3) The samples with gold anomaly are randomly picked up according to the ratios of each interval, whereas samples with other element anomaly are orderly picked up from those samples characterized by higher TAN, HAN and LAN values. (4) If necessary, supplementary samples are randomly picked up from certain analytical batches. The MIEGSP takes into account, not only geochemical anomalies and experimental analysis anomalies, but also supplements of non-anomalies and analytical batch balance sampling. It has resolved the paradox between limited sample amount and the lack of sampling representativeness. An EXCEL program for anomaly sampling based on the Macros (VBA) was developed. The MIEGSP has been successfully applied to routine analysis of geochemical reconnaissance survey samples.

     

/

返回文章
返回