Talk:Level 2 data (velocity profilers): Difference between revisions
From Atomix
Created page with "~~~~ The description of the segmented data as being detrended implies a distinction from the standard level 2 data. The “standard” level 2 data should also be detrended -..." |
No edit summary |
||
| Line 1: | Line 1: | ||
[[User:Brian scannell|Brian scannell]] ([[User talk:Brian scannell|talk]]) 15:51, 28 December 2021 (CET) The description of the segmented data as being detrended implies a distinction from the standard level 2 data. The “standard” level 2 data should also be detrended - but across the burst rather than a segment. Makes me wonder whether having both burst and segment adds anything other than possible confusion? Perhaps omit BURST_NUMBER at level 1 and introduce SEGMENT_NUMBER as a required variable at level 2? | [[User:Brian scannell|Brian scannell]] ([[User talk:Brian scannell|talk]]) 15:51, 28 December 2021 (CET) The description of the segmented data as being detrended implies a distinction from the standard level 2 data. The “standard” level 2 data should also be detrended - but across the burst rather than a segment. Makes me wonder whether having both burst and segment adds anything other than possible confusion? Perhaps omit BURST_NUMBER at level 1 and introduce SEGMENT_NUMBER as a required variable at level 2? Having SAMPLE_NUMBER (or possibly PROFILE_NUMBER) as a level 1 variable (dimension TIME) replicates TIME as a unique identifier for each set of observations (dimensions R_DIST, N_BEAM) but as an integer variable it would be easier to work with when defining bursts / segments. | ||
Revision as of 15:01, 28 December 2021
Brian scannell (talk) 15:51, 28 December 2021 (CET) The description of the segmented data as being detrended implies a distinction from the standard level 2 data. The “standard” level 2 data should also be detrended - but across the burst rather than a segment. Makes me wonder whether having both burst and segment adds anything other than possible confusion? Perhaps omit BURST_NUMBER at level 1 and introduce SEGMENT_NUMBER as a required variable at level 2? Having SAMPLE_NUMBER (or possibly PROFILE_NUMBER) as a level 1 variable (dimension TIME) replicates TIME as a unique identifier for each set of observations (dimensions R_DIST, N_BEAM) but as an integer variable it would be easier to work with when defining bursts / segments.
