Chapter 12 — Data Evaluation and Interpretation¶
12.1 Evaluation turns data into understanding¶
Collecting data is not monitoring; evaluating it is. Evaluation asks whether measured behavior matches expected behavior (Chapter 1) and what a deviation means.
12.2 Baselines and expected behavior¶
Expected behavior is established by:
- Design analyses (stability, seepage, deformation).
- The early-operation baseline — the first years of readings under known conditions.
- Empirical norms for similar facilities and instruments.
Deviations are judged against these, not against arbitrary numbers.
12.3 Plots that reveal signal¶
- Time-series of each parameter against pond level, rainfall, and season.
- Cross-plots (e.g., crest settlement vs. piezometric head).
- Profiles (pore pressure with depth; deformation with distance).
- Cumulative vs. rate plots to distinguish steady from accelerating movement.
12.4 Separating signal from noise¶
All instruments have noise. The interpreter must distinguish:
- Real trend (monotonic change with a physical cause).
- Seasonal/operational variation (pond cycle, temperature, deposition).
- Instrument artifact (drift, clogging, electrical fault).
Correlation with independent parameters (does movement follow pore pressure?) is the strongest test of a real signal.
12.5 The role of satellite InSAR in evaluation¶
Wide-area displacement maps from InSAR reveal movement that point instruments miss — and confirm that point instruments are representative. Anomalous InSAR zones warrant targeted ground investigation.
Rate of change beats magnitude
Acceleration — a change in the rate of movement or pore-pressure rise — is often the earliest and most reliable warning. Watch derivatives, not just values.
12.6 Key takeaways¶
- Evaluation compares measured to expected behavior.
- Build baselines from design, early operation, and norms.
- Use time-series, cross-plots, and profiles to find signal.
- Watch rates of change; acceleration is the earliest warning.