smwoodman
commented
2 weeks ago Big picture question:
- [ ] Sam is still working with the developers, but currently pyglider only let's you process data while interpolating to a given sensor. Do we need an ‘L0-raw’ timeseries file with all variables and no interpolation? Or are we ok with one file that has all of the timestamps/variables, with the following interpolations:
- eng vars interpolated to m_depth
- ctd vars interpolated to sci_water_temp
- oxygen vars interpolated to sci_oxy4_oxygen
- ecopuck/optics vars interpolated to sci_flbbcd_chlor_units
- lat/lon interpolated for the whole time series
More specific questions:
- [ ] In SOCIB, L0 means no interpolation/processing, L1 means interpolation/removing bad values/etc, and L2 means gridded, correct?
- [ ] For our base timeseries, do we want time values in seconds or nanoseconds?
- [ ] Decide on values used when calculating profiles: profile_filt_time (default 100) = time in seconds over which to smooth the pressure time series for finding up and down profiles; profile_min_time (default 300) = minimum time to consider a profile an actual profile (seconds)
- [ ] What other filtering should automatically happen? For instance, removing all pressure sensor values less than 0, removing salinity values outside of some range, ...
Note from past convos with CR: we need both 1m and 5m profiles. 1m for the oceanography and optics, 5m for the acoustics. We should also try and use the same depth (1010m) and surface cutoff 5m as SOCIB, so we remove bad data at the surface and always have a boundary at the bottom. (it will also make our arrays the same size in the future x by 1010 or 201 bins).
This issue is to track decisions made around choices for base glider processing. For instance, what/when to interpolate, what variables need to be included in the output, etc. These questions will mostly need to be answered by Christian. They are checked when they have been answered.