|Title||Geochemical data quality control for soil, till and lake and stream sediment samples|
|Author||McCurdy, M W; Garrett, R G|
|Source||Geological Survey of Canada, Open File 7944, 2016, 40 pages, https://doi.org/10.4095/297562 (Open Access)|
|Links||Canadian Database of Geochemical Surveys, downloadable
|Links||Banque de données de levés géochimiques du Canada,
|Publisher||Natural Resources Canada|
|Subjects||geochemistry; geochemical analyses; geochemical interpretations; geochemical surveys; data collections; field data methods; stream sediment geochemistry; stream water geochemistry; soil geochemistry;
till geochemistry; lake sediment geochemistry; lake water geochemistry|
|Program||Hudson/Ungava, Northeastern Quebec-Labrador, surficial geology, GEM2: Geo-mapping for Energy and Minerals|
|Released||2016 01 28|
Geochemical data for lake and stream sediments, soils and tills received from commercial and government laboratories need to be evaluated for quality in terms of accuracy and
precision before release. Quality control (QC) of National Geochemical Reconnaissance data is carried out using Microsoft Excel® with the Open Source R Project for Statistical Computing software (R Core Team, 2015) and 'rgr', 'The GSC Applied
Geochemistry EDA Package' (Garrett 2013a, 2015a), that sits on top of R and provides the functionality required for many QA/QC tasks. Routines in 'rgr' are used to evaluate small (<20 samples) using graphical methods and larger batches using both
graphical and statistical methods. Accuracy is estimated by inserting Certified Reference Materials (CRM) with accepted mean concentrations determined by different analytical methods. Precision is estimated using analytical duplicate pairs. For large
batches of samples (>30 pairs of observations) an analysis of variance (ANOVA) is carried out using field duplicate pairs.
This document represents the first step in a process with the goal of building a consistent methodology for evaluating the
quality of geochemical analytical data before they are released to the public. For the most part, the procedures described below, especially the construction of tables for Accuracy, Precision and ANOVA, are both time-consuming and tedious. However,
there are a number of open source programs available that can be used in conjunction with R and Microsoft Excel® that could expedite the generation of statistics and the process of filling in the rows and columns of the tables. These include RExcel,
an add-in to Excel that allows the use of R as a 'helper application' for Excel (Heiberger and Neuwirth, 2009). RStudio and Rcmdr are menu interfaces for R, also free and Open Source that could be integrated with R and Excel to create a system for
evaluating the quality of geochemical data in a timely and efficient manner.
The reader is advised that the GSC has never had a single institutional approach to geochemical quality control. The procedures described herein have been developed by
one group at one location (601 Booth Street, Ottawa). Other groups have instituted other procedures.
|Summary||(Plain Language Summary, not published)|
This open file provides graphical and statistical procedures to estimate the accuracy and precision of lake and stream sediment geochemical data returned
from commercial and government laboratories. These procedures are used to evaluate data for the National Geochemical Reconnaissance program at the Geological Survey of Canada.