echopype 0.1.2

Enhancing interoperability and scalability in analyzing water column sonar data.

The goal of echopype is to improve the interoperability and scalability in processing water column sonar data. These data are commonly used for inferring information about mid-trophic organisms in the ocean, such as fish and krill.

Functionality

echopype include tools for:

• Converting different manufacturer-specifc data files into an interoperable netCDF format.

• Processing large volumes of sonar data in local or cloud storage by leveraging Python distributed computing libraries.

The current version supports file conversion for the .raw data files collected by the SimRad EK60 echosounder. Conversion for other types of data files, including the .01A files from AZFP echosounder, the .raw files from the SimRad broadband EK80 echosounder, and the raw beam data from ADCP (Acoustic Doppler Current Profiler) will be added in future releases.

Installation

To install echopype, do the following in your terminal:

$ pip install echopype

Using echopype

File conversion

To batch convert .raw files to the interoperable netCDF format in the terminal, do:

$ echopype_converter -s ek60 data/*.raw

This will generate corresponding .nc files with the same leading filename as the original .raw files in the same directory.

To use the EK60 data converter in a Python session, you can do:

# import as part of a submodule
from echopype.convert import ConvertEK60
data_tmp = ConvertEK60('FILENAME.raw')
data_tmp.raw2nc()

Or:

# import the full module
import echopype as ep
data_tmp = ep.convert.ConvertEK60('FILENAME.raw')
data_tmp.raw2nc()

Same as in the command line case, this will generate a FILENAME.nc in the same directory as FILENAME.raw.

The ConvertEK60 instance contains all the data unpacked from the .raw file, so it is a good idea to clear it from memory once done with conversion.

Data analysis

The data analysis functionalites of echopype is being developed actively. echopype currently supports:

• calibration and echo-integration to obtain volume backscattering strength (Sv) from the power data collected by EK60.

• simple noise removal by suppressing data points below an adaptively estimated noise floor [1].

• binning and averaging to obtain mean volume backscattering strength (MVBS) from the calibrated data.

The steps of performing these analysis are summarized below:

from echopype.model import EchoData
data = EchoData('FILENAME.nc')
data.calibrate()  # Calibration and echo-integration
data.remove_noise(save=True)  # Save denoised Sv to FILENAME_Sv_clean.nc
data.get_MVBS(save=True)  # Save MVBS to FILENAME_MVBS.nc

Note that by default, method calibrate save the calibrated volume backscattering strength (Sv) to FILENAME_Sv.nc, while method remove_noise and get_MVBS by default do not generate new files. The computation results from these two methods can be accessed from data.Sv_clean and data.MVBS as xarray DataSets. All outputs are xarray DataSets with proper dimension labeling.

License

echopype is licensed under the open source Apache 2.0 license.

This project is lead by Wu-Jung Lee (@leewujung). Other contributors include:

• Valentina Staneva (@valentina-s)

• Marian Peña (@marianpena)

• Mark Langhirt (@bnwkeys)

• Erin Labrecque (@erinann)

• Emma Ozanich (@emma-ozanich)

• Aaron Marburg (@amarburg)

References

[1]

De Robertis and Higginbottoms (2007) A post-processing technique to estimate the signal-to-noise ratio and remove echosounder background noise. ICES J. Mar. Sci. 64(6): 1282–1291.

---

Copyright (c) 2018–, Wu-Jung Lee, Applied Physics Laboratory, University of Washington.