Hi,
When I upload 1794041449-GIR.fit to Intervals.icu, it shows approximately 500 m of elevation gain for a run of just 1.23 km lasting 7 minutes 13 seconds.
An analysis of the FIT file identified abnormal altitude readings during the first 10 seconds:
| Time from start | Recorded altitude |
|---|---|
| 0–1 s | 590.8 m |
| 2–3 s | 0 m |
| 4 s | 539.8 m |
| 6 s | 532.4 m |
| 7 s | 518.0 m |
| 9 s | 496.6 m |
| 10 s | 489.4 m |
After that, the altitude readings become consistent, with small, gradual changes. Summing the altitude changes from the 10-second mark onwards gives approximately 10.2 m of ascent and 15.2 m of descent. These are estimates for the remaining activity, rather than verified totals for the entire run.
The source file clearly contains faulty initial readings, but Intervals.icu appears to include the resulting spike in its elevation gain calculation.
Could you investigate whether these implausible altitude jumps can be detected and excluded during import, or suggest an existing way to correct the elevation data?
The issue can be reproduced by importing the original FIT file, which I can provide for troubleshooting.
Thanks!