Data-Quality Measures of Stakeholder-Implemented Watershed-Monitoring Programs
by Adrienne I. Greve
Available from the U.S. Geological Survey, Branch of Information Services, Box 25286, Denver Federal Center, Denver, CO 80225, USGS Open-File Report 02–141, 19 p., 5 figs.
This document also is available in pdf format: 
OFR 02–141.pdf (909K)
(Requires Adobe
Acrobat Reader)
Abstract
Community-based watershed groups, many of which collect environmental data, have steadily increased in number over the last decade. The data generated by these programs are often underutilized due to uncertainty in the quality of data produced. The incorporation of data-quality measures into stakeholder monitoring programs lends statistical validity to data.
Data-quality measures are divided into three steps: quality assurance, quality control, and quality assessment. The quality-assurance step attempts to control sources of error that cannot be directly quantified. This step is part of the design phase of a monitoring program and includes clearly defined, quantifiable objectives, sampling sites that meet the objectives, standardized protocols for sample collection, and standardized laboratory methods. Quality control (QC) is the collection of samples to assess the magnitude of error in a data set due to sampling, processing, transport, and analysis. In order to design a QC sampling program, a series of issues needs to be considered: (1) potential sources of error, (2) the type of QC samples, (3) inference space, (4) the number of QC samples, and (5) the distribution of the QC samples. Quality assessment is the process of evaluating quality-assurance measures and analyzing the QC data in order to interpret the environmental data. Quality assessment has two parts: one that is conducted on an ongoing basis as the monitoring program is running, and one that is conducted during the analysis of environmental data.
The discussion of the data-quality measures is followed by an example of their application to a monitoring program in the Big Thompson River watershed of northern Colorado.
Table of Contents
Glossary
Abstract
Introduction
Purpose and Scope
Acknowledgments
Quality Assurance
Possible Quality-Assurance Approaches for Stakeholder Groups
Quality Control
Quality-Control Sample Design
Determining the Potential Sources of Error
Type of Quality-Control Samples Needed
Determining the Inference Space for Quality-Control Samples
Number of Quality-Control Samples Needed
Distribution of Quality-Control Samples within an Inference Space
Quality Assessment
Ongoing Quality-Assessment Measures
Using Quality-Assessment Measures during Environmental Data Analysis
Estimating Variability by Using Field-Replicate Quality-Control Samples
Interpreting Environmental Data by Using Field-Replicate Quality-Control Samples
Estimating Bias by Using Field-Blank Quality-Control Samples
Interpreting Environmental Data by Using an Estimate of Bias
Estimating Matrix Interaction and Sample Degradation with Field-Spike Data
Interpreting Environmental Data by Using Spike-Recovery Data
A Data-Quality Program in the Big Thompson River Watershed
Quality Assurance for the Big Thompson Watershed Forum
Quality Control for the Big Thompson Watershed Forum
Types of Basic Quality-Control Samples to be Collected
Types of Topical Quality-Control Samples to be Collected
Inference Space
The Number of Quality-Control Samples to be Collected
Distribution of Quality-Control Samples
Summary and Conclusions
References
