CSM Data Collection & Analysis 2 — Questions and Answers
Question 1: Which statistical method is most appropriate for identifying the relationship between impervious surface percentage and peak runoff volume across multiple watersheds?
- Chi-square test
- Linear regression analysis (Correct answer)
- Mann-Whitney U test
- Kruskal-Wallis test
Correct answer: Linear regression analysis
Linear regression analysis quantifies the linear relationship between a continuous predictor (impervious cover %) and a continuous outcome (peak runoff), making it the most appropriate tool here.
Question 2: A stormwater manager receives turbidity readings of 5, 6, 200, 7, and 5 NTU from five consecutive samples. The value of 200 NTU is most likely a(n):
- Representative peak event value
- Outlier requiring investigation (Correct answer)
- Calibration standard result
- Normal diurnal variation
Correct answer: Outlier requiring investigation
A single value orders of magnitude above the rest of the dataset is a statistical outlier and should be investigated for sampling error, equipment malfunction, or a true transient event.
Question 3: When designing a stormwater monitoring network, the concept of 'spatial representativeness' requires that sampling locations:
- Are placed exclusively at the watershed outlet
- Capture the range of land uses and sub-watershed conditions (Correct answer)
- Are spaced exactly 500 meters apart
- Only sample during storm events larger than the 10-year return period
Correct answer: Capture the range of land uses and sub-watershed conditions
Spatial representativeness means monitoring locations are chosen to reflect the variety of conditions across a watershed so data generalizes to the whole system.
Question 4: A flow-weighted composite sample is collected by:
- Taking equal-volume aliquots at fixed time intervals
- Taking aliquots proportional to instantaneous stream flow at each interval (Correct answer)
- Collecting only during baseflow conditions
- Combining samples from multiple watershed outfalls in equal parts
Correct answer: Taking aliquots proportional to instantaneous stream flow at each interval
Flow-weighted compositing adds aliquot volumes proportional to flow so the composite reflects actual pollutant mass loading rather than time-averaged concentration.
Question 5: Which data quality indicator (DQI) addresses whether a measurement is close to the true or accepted reference value?
- Precision
- Accuracy (Correct answer)
- Completeness
- Representativeness
Correct answer: Accuracy
Accuracy (or bias) measures how close a result is to the true value, often assessed using certified reference standards or spikes.
Question 6: The US EPA's NPDES Phase II permit requires small MS4s to track the effectiveness of BMPs. Which metric best quantifies pollutant load reduction?
- Event mean concentration (EMC) alone
- Annual pollutant load (mass/year) before and after BMP installation (Correct answer)
- Grab sample turbidity at BMP inlet only
- Percentage of impervious area within the MS4 boundary
Correct answer: Annual pollutant load (mass/year) before and after BMP installation
Annual pollutant load comparisons (concentration × flow volume × time) before and after BMP installation directly measure mass reduction, which is the relevant metric for permit compliance.
Question 7: A stormwater dataset has a highly skewed distribution of pollutant concentrations. Which measure of central tendency is most appropriate to report?
- Arithmetic mean
- Median (Correct answer)
- Mode
- Range
Correct answer: Median
The median is resistant to the influence of extreme high values that are common in skewed stormwater pollutant datasets, making it a more representative central tendency measure than the mean.
Which statistical method is most appropriate for identifying the relationship between impervious surface percentage and peak runoff volume across multiple watersheds?