Gradient method
Tsurutani and Smith (1979)
Our criteria were applied to 1-min averages of the three field components representing an average vector \(B_i\). The vector field change between this vector and the vector averaged 3 min earlier was computed, i.e., \(ΔB = B_i - B_{i-3}\), as were the three field magnitude \(|B_i|\), \(|B_{i-3}|\), and \(|ΔB|\). Discontinuities were selected by requiring that the magnitude of the vector change equal or exceed one half the larger of \(|B_i|\) and \(|B_{i-3}|\), which we call \(B_L\); that is, we require that \(|ΔB|> B_L/2\).
detect_gradient
detect_gradient (df:polars.lazyframe.frame.LazyFrame, cols:list[str], time:str='time', avg_interval=datetime.timedelta(seconds=60), window=datetime.timedelta(seconds=180))
Identifies discontinuities in the averaged vector field and specified criteria: |ΔB| > max(|B_i|, |B_{i-window}|) / 2
| Type | Default | Details | |
|---|---|---|---|
| df | LazyFrame | Dataframe containing the raw vector components. | |
| cols | list | ||
| time | str | time | Column name for time. |
| avg_interval | timedelta | 0:01:00 | Time interval for averaging (e.g., ‘1m’ for 1 minute). |
| window | timedelta | 0:03:00 | Time interval to look back for computing ΔB. |
References
Tsurutani, Bruce T., and Edward J. Smith. 1979. “Interplanetary Discontinuities: Temporal Variations and the Radial Gradient from 1 to 8.5 AU.” Journal of Geophysical Research: Space Physics 84 (A6): 2773–87. https://doi.org/10.1029/JA084iA06p02773.