Transcript Mod13-B QA/QC for Environmental Measurement

QA/QC FOR
ENVIRONMENTAL
MEASUREMENT
Unit 4: Module 13, Lecture 2
Objectives
 Introduce the why and how of Quality Control
 Analysis of natural systems
 Why do we need QC?
 Introduce Data Quality Objectives (DQOs)
 How do we evaluate quality of data ?
 Emphasize the PARCC parameters
 QC sample(s) applicable for each key parameter
 QC sample collection and evaluation methods
 Statistical calculation of percussion
 Determination of accuracy and bias
 Introduce Quality Assurance Project Plans
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s2
Quality Control
 What is Quality Control (QC)?
 The overall system of technical activities
designed to measure quality and limit error in a
product or service.
 A QC program manages quality so that data
meets the needs of the user as expressed in a
Quality Assurance Program Plan (QAPP).
- US EPA (1996)
QC is used to provide QUALITY DATA
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s3
QC for environmental measurement
 Evaluation of a natural system:
 Collect environmental samples
 Specified matrix – medium to be tested (e.g.
soil, surface water, etc.)
 Specified analytes – property or substance
to be measured (e.g. pH, dissolved oxygen,
bacteria, heavy metals)
http://ma.water.usgs.gov/CapeCodToxi
cs/photo-gallery/wq-sampling.htm
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s4
QC for environmental measurement
 QC is particularly critical in field data collection
 often the most costly aspect of a project
 data is never reproducible under the exact same
condition or setting
http://climchange.cr.usgs.gov/info/lacs/wate
rsampling.htm
sechi readings
Developed by: Zwiebel, Filbin
field filtration
Updated: June 14, 2005
http://www.fe.doe.gov/techline/tl_
hydrates_oregon.shtml
logging sea cores
U5-m13b-s5
QC for environmental measurement
 Natural systems are inherently variable
 Variability of lakes vs. streams vs. estuaries
 Changes in temperature, sunlight, flow, sediment
load and inhabitants
 Human introduction of error
http://pubs.usgs.gov/fs/fs-0058-99
http://www.nrcs.usda.gov/programs/cta/ctasummary.html
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s6
QC for environmental measurement
 Why do we need quality control?
 To prevent errors from happening
 To identify and correct errors that have taken
place
QC is used to PREVENT and CORRECT ERRORS
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s7
QC for environmental measurement
 QC systems are used to:
 Provide constant checks on sensitivity and
accuracy of instruments.
 Maintain instrument calibration and accurate
response.
 Provide real-time monitoring of instrument
performance.
 Monitor long-term performance of measurement
and analytical systems (Control Charts) and
correct biases when detected.
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s8
QC for environmental measurement
 Data Quality Objectives (DQOs):
 Unique to the goals of each environmental
evaluation
 Address usability of data to the data user(s)
 Those who will be evaluating or employing data
results
 Specify quality and quantity of data needed
 Include indicators such as precision, accuracy,
representativeness, comparability, and
completeness (PARCC); and sensitivity.
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s9
QC for environmental measurement
 The PARCC parameters help evaluate sources
of variability and error
 Precision
 Accuracy
 Representativeness
 Completeness
 Comparability
“PARCC” parameters increase the level of
confidence in our data
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s10
QC for environmental measurement
 Sensitivity
 Ability to discriminate between measurement
responses
 Detection limit
 Lowest concentration accurately detectable
 Instrument detection limit
 Method detection limit (MDL)
 Measurement range
 Extent of reliability for instrument readings
 Provided by the manufacturer
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s11
Quality control methods: QC samples
 Greater that 50% of all errors
found in environmental
analysis can be directly
attributed to incorrect sampling
 Contamination
 Improper preservation
 Lacking representativeness
 Quality control (QC) samples
are a way to evaluate the
PARCC parameters.
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s12
Quality control methods: QC samples
QC sample types include:
 field blank
 equipment or rinsate blank
 duplicate/replicate samples
 spiked samples
 split samples
 blind samples
http://ma.water.usgs.gov/CapeCodToxics/photo-gallery/wq-sampling.htm
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s13
Quality control methods: QC samples
 Field blank sample collection
 In the field, using a sample container supplied by
the analytical laboratory, collect a sample of
analyte free water (e.g. distilled water)
 Use preservative if required for other samples
 Treat the sample the same as all other samples
collected during the designated sampling period
 Submit the blank for analysis with the other
samples from that field operation.
 Field blanks determine representativeness
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s14
Quality control methods: QC samples
 Equipment or rinsate blank collection
 Rinse the equipment to be used in sampling with
distilled water immediately prior to collecting the
sample
 Treat the sample the same as all others, use
preservative if required for analysis of the batch
 Submit the collected rinsate for analysis, along
with samples from that sample batch
 Rinsate blanks determine representativeness
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s15
Quality control methods: QC samples
 Duplicate or Replicate sample collection
 Two separate samples are collected at the same
time, location, and using the same method
 The samples are to be carried through all
assessment and analytical procedures in an
identical but independent manner
 More that two duplicate samples are called
replicate samples.
 Replicates determine representativeness
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s16
QC methods: Representativeness
 Representativeness  extent to which measurements actually represent
the true environmental condition or population at
the time a sample was collected.
 Representative data should result in repeatable
data
 Does this
represent this??

http://pubs.usgs.gov/fs/fs-0058-99
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s17
Quality control methods: QC samples
 Split and blind sample collection
 A sample is collected and mixed thoroughly
 The sample is divided equally into 2 or more
sub-samples and submitted to different analysts
or laboratories.
 Field split
 Lab split
 Blinds - submitted without analysts knowledge
 Split and blind samples determine precision
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s18
Quality control methods: QC samples
 Spiked sample preparation
 A known concentration of the analyte is added to
the sample
 Field preparation
 Lab preparation
 The sample is treated the same as others for all
assessment and analytical procedures
 Spiked samples determine accuracy
 % recovery of the spiked material is used to
calculate accuracy
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s19
Quality control methods: QC Samples
 Precision  degree of agreement
between repeated
measurements of the
same characteristic
 can be biased –
meaning a consistent
error may exist in the
results
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s20
Key concepts of QA/QC: Precision
 Precision –
 degree of agreement
target images
between results
 Statistical Precision  standard deviation,
Adapted from Ratti and Garton (1994)
or relative percent
difference from the
mean value
Mean Value
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s21
Key concepts of QA/QC: Precision
How to quantify precision:
1. Determine the mean result of the data (the
average value for the data)

the arithmetic mean will usually work.
To determine arithmetic mean:
1. add up the value of each data point
2. divide by the total number of points “n”
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
Mean Value
U5-m13b-s22
Key concepts of QA/QC: Precision
How to quantify precision:
2. Determine the first and second standard
deviation (SD).


SD1 = approximately 68% of the data points
included on either side of the mean
SD2 = approximately 95% of the data points
included on either side of the mean
Mean Value
SD1
SD1
SD2
SD2
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s23
Key concepts of QA/QC: Precision
 The lower diagrams show ‘scatter’
around the mean
 The SD quantifies the degree of
scatter (or spread of data)
 Less scatter = smaller SD value
and grater precision (target 1)
Mean Value
(18.48)
Adapted from Ratti and Garton
(1994)
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s24
Key concepts of QA/QC: Precision
 Improbable Data
 Data values outside the 95th (2 SD) interval (below)
 These are improbable
-2.0
Developed by: Zwiebel, Filbin
-1.5
2.0
-1.0
1.0 -0.5
0.0
0
Updated: June 14, 2005
0.5
1.0 1.0
1.5
2.0
2.0
U5-m13b-s25
Key concepts of QA/QC: Precision
 Below example: The mean value 18.480C
 The standard deviation SD is 2.340C
 The precision value is expressed 18.480C +/- 2.340C
Mean Value (18.48)
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s26
Key concepts of QA/QC: Accuracy
 accuracy = (average
value) – (true value)
 precision represents
repeatability
 bias represents
amount of error
 low bias and high
precision = statistical
accuracy
http://www.epa.gov/owow/monitoring/volunteer/qappexec.html
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s27
Key concepts of QA/QC: accuracy & bias
 Determine the accuracy and bias of this data:
Example Data Collected - pH 7.0 Standard
Group 1
Group 2
Group 3
Group 4
7.5
7.2
6.5
7.0
7.4
6.8
7.2
7.4
6.7
7.3
6.8
7.2
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s28
Key concepts of QA/QC: Comparability
 Comparability  the extent to which data generated by different
methods and data sets are comparable
 Variations in the sensitivity of the instruments
and analysis used to collect and assess data will
have an effect upon comparability with other
data sets.
 Will similar data from
these instruments be
Comparable ?? 
Hach DR2400 portable spectrophotometer
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s29
Key concepts of QA/QC: Completeness
 Completeness  % comparison between the amount of data
intended to be collected vs. actual amount of
valid (usable) data collected.
 In the QAPP design – do the goals of the plan
meet assessment needs?
 Will sufficient data be collected?
Would this give usable data ??
Developed by: Zwiebel, Filbin

Updated: June 14, 2005
U5-m13b-s30
Key concepts of QA/QC: Completeness
 Sample design
 Will samples
collected at an out
flow characterize
conditions in the
entire lake?
 Statistically relevant
number of data points
 Will analysis in ppm
address analytes
toxic at ppb?
Developed by: Zwiebel, Filbin
 Valid data
 Would data be
sufficient if high
humidity resulted in
“error” readings?
 Is data valid if the
readings are outside
the measurement
range of the
instrument?
Updated: June 14, 2005
U5-m13b-s31
Review: Quality Assurance Project Plans
 The QAPP is a
project-specific QA
document.
 The QAPP outlines
the QC measures to
be taken for the
project.
Developed by: Zwiebel, Filbin
 QAPP guides:
 the selection of
parameters and
procedures
 data management
and analysis
 steps taken to
determine the
validity of specific
sampling or analysis
procedures
Updated: June 14, 2005
U5-m13b-s32
Review: Elements of a QAPP
 The QAPP governs work conducted in the field,
laboratory, and the office.
 The QAPP consists of 24 elements generally
grouped into four project areas:
 Project management (office)
 Measurement and data acquisition (field and lab)
 Assessment and oversight (field, lab, and office)
 Data validation and usability (field, lab, and
office)
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s33
References
 EPA 1996, Environmental Protection Agency Volunteer
Monitor’s Guide to: Quality Assurance Project Plans. 1996.
EPA 841-B-96-003, Sep 1996, U.S. EPA, Office of
Wetlands, Washington, D.C. 20460, USA
http://www.epa.gov/owowwtr1/monitoring/volunteer/qappex
ec.htm
 EPA 1994, Environmental Protection Agency Requirements
for Quality Assurance Project Plans for Environmental Data
Operations. EPA QA/R-5, August 1994). U.S. EPA,
Washington, D.C. 20460, USA
 Ratti, J.T., and E.O. Garton. 1994. Research and
experimental design. pages 1-23 in T.A. Bookhout, editor.
Research and management techniques for wildlife and
habitats. The Wildlife Society, Bethesda, Md.
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s34
Developed by: Zwiebel, Filbin
Updated: June 14, 2005
U5-m13b-s35