search

OSOceanAcoustics / echopype

Enabling interoperability and scalability in ocean sonar data analysis
https://echopype.readthedocs.io/
Apache License 2.0
99 stars 76 forks source link

Re-organize AD2CP data into different `Beam_groupX` groups #711

Open leewujung opened 2 years ago

leewujung commented 2 years ago

The v0.6.0 change of netCDF data structure did not include the Nortek Signature ADCP (AD2CP) data, because the data is of substantially different format compared to the EK and AZFP echosounders. But now with the experience (and struggle!) with forcing data from the other models into a more compliant form for SONAR-netCDF4 ver.1 convention, I think we can take this step in the upcoming releases.

The main thing is that currently all backscatter and velocity data are saved in a single Beam group, and because of it likely result into a large number of NaNs due to the different sampling time and resolution/range setting in the different "modes" the AD2CP supports. In v0.6.0 we moved the original Beam (and Beam_power) group to be Sonar/Beam_group1 (and Sonar/Beam_group2) for the EK and AZFP echosounders. Taking the same approach, we can save the data from different AD2CP modes into different Beam_groupX so that it would be straightforward to find and use.

This issue will serve as the drafting and discussion space for this change.

leewujung commented 2 years ago

Proposed variable movements and attribute elaboration in #719.

leewujung commented 2 years ago

Hey @imranmaj : once you have something in place for the changes we discussed yesterday please push up a draft PR so that I (and @emiliom ) can review the changes in smaller chunks. Thanks!

leewujung commented 2 years ago

The majority of variable movements are completed in #731. I'll keep this open but defer the rest to v0.6.2.