2 question you need to ask when composing a sampling approach
- What kinds of samples are needed
2. How many of them are needed
Additional questions for sampling approach
- how will will the results be used?
- where to collect
- what containers
- how to preserve
- what preparations are necessary
- how to analyze(instruments,detectors, analytical methods)
- to to evaluate data
What are the four principles for sampling plan
- environmental samples must be representative of the portion of the environment being investigated
- procedures for sampling and analysis influence each other, plans for sampling and analysis are codependent
- Quality control samples must be representative of the environmental samples being analyzed
- QCs used to assess amount of bias and imprecision in data from analysis of samples
What is the order or error pyramid
- data handing (top of pyramid)
- Analysis
- sample prep
- sample handling
- field sampling
- pollutant distribution (bottom of pyramid)
What is a DQOs
Data quality objectives are a process (or set of terms) that outline the overall uncertainty that can be tolerated in the project. Qualitative or quantitative.
The point is to make a corrective action if DQOs aren’t met
Work on principle 2
What are the 2 major components of DQO process
- very specifically state your question(a)
2. very specifically state amount of uncertainty your willing to tolerate
7 steps of DQOs utilization
- Problem
- Identity decision required
- identify inputs to decision
- determine the boundaries of study
- Develop a decision rule
- develop uncertainty constraints
- optimize the design for obtaining data
Types of sampling techniques
- random sampling
- no bias
- systematic random sampling
3.judgmental sampling
Explain random sampling
Every sample of the population has an equal chance of being selected
-very expensive, unbiased, assumes samples are independent but spacially correlated, samples are farther apart then the range of correlation
Explain what semi variogram is
A tool used to determine the number of samples required for a program to be considered statistically random.
-shows how distance between samples increases randomness
(Dist can be spatial or temporal)
Explain systematic random sampling
Divides population it to homogeneous groups to reduce variance.
-groups divided into cells by Cartesian grid
-slightly more bias, used for contaminated land sites, drill wells, biological sampling
Explain judgemental sampling
Most cost effective, most bias, based on prior history and visual site assessment,doesn’t factor in pollutant behaviour or migration
Factors for choosing a sampling plan
1type of pollutant
2type of sampling device
3end use of data
