Why do saved output times change estimated storm peak intensity?
Coarse output intervals can miss the strongest simulated updraft even when the model integrates every event internally. Diagnose the sampling effect by thinning output from one run.
Start with How do you track an updraft centroid in model output?.
| Minute | 0 | 1 | 2 | 3 | 4 |
|---|---|---|---|---|---|
| Updraft speed (m/s) | 10 | 18 | 31 | 19 | 12 |
Set up the experiment
Use the same five minute-by-minute values in both samples. Predict what happens when only minutes 0 and 4 are retained. The full series peaks at 31 m/s at minute 2. The sparse output reports a maximum of 12 m/s, 19 m/s below that sampled peak. Neither record establishes the maximum between its own output times.
Practice questions
- Save at minutes 0, 2 and 4. What maximum survives?
- Save at minutes 1 and 3. What maximum survives?
- Does this experiment change the numerical solution?
Worked answers
Answer 1
31 m/s, because the sampled peak is retained.
Answer 2
19 m/s, which misses the sampled peak by 12 m/s.
Answer 3
No. Thinning the saved values changes the available diagnostics while leaving the original run unchanged.
- 31 m/s, because the sampled peak is retained.
- 19 m/s, which misses the sampled peak by 12 m/s.
- No. Thinning the saved values changes the available diagnostics while leaving the original run unchanged.