Project 1 - Process & Archive Soundscape Data

Goal and Target Outcome

We will facilitate cross-agency, region, and platform soundscape comparisons by quantifying soundscape metrics for priority datasets and generating decision support tools that integrate soundscape visualizations with multiple data types. This will allow for systematic identification, archival, and processing of soundscape metrics for priority PAM-SI datasets to build capacity and consistency across programs.

Summary

This project will add soundscape metrics for priority PAM SI datasets alongside audio files and other metrics to the NCEI Passive Acoustics archive. The soundscape metrics will be used to develop and demonstrate soundscape interpretation and visualization tools being developed as part of the SI. The datasets will also facilitate comparison of data sampled by multiple recording technologies and projects that are unique to select FMCs. A technical guide for processing and archiving PAM data will also be prepared as part of this project. This project will work with PAM SI cloud computing projects to operationalize soundscape metric processing software in the cloud environment. Streamlining and documenting methods to process and archive soundscape metrics supports PAM SI objectives to advance PAM data analysis and integration approaches, and advance archival cloud capabilities.

Working Group

Establish data processing group, Google doc folder, regular check-in schedule.

Identify priority datasets

In coordination with Carrie Wall Bell and NCEI Archiving group decide on soundscape priority datasets for each FMC and flag in the overall PAM SI priority dataset table.

Prepare analyses

For priority soundscape datasets: Prepare metadata, share QA/QC protocol/tools, prepare and submit audio data and soundscape metrics (if already processed locally) to NCEI for archiving.

Test in pyPAM

Coordinate with PAM SI Cloud team to test pyPAM cloud processing from NCEI GCP on priority NRS audio data sets.

Establish software standards

Identify standard software options available for use and test Noise Reference Station (NRS) dataset results from multiple software outputs.

Compare soundscape software options

Provide soundscape software options available (MANTA, pyPAM, updated Triton toolbox, etc.) with comparison documentation for example datasets (NRS or others if necessary).

Generate Hybrid Milli-Decade (HMD) output

FMCs will conduct HMD processing either locally or through the PAM SI Cloud computing and save to NCEI or internal PAM SI cloud environment.

NCEI archive data products

All HMD outputs will be processed and sent to NCEI archive for level 2-3 data products.

Facilitate more complex soundscape analyses

All priority soundscape datasets and HMD level 1 products will be cloud archived at NCEI for use in the Soundscape interpretation and visualization project.

Create supporting documents

Develop technical reports and tutorial videos for HMD metric processing, minimum QA/QC, and visualization routines developed in coordination with SoundCoop efforts, and PyPAM and PAMscapes software development efforts.

Support comparison across platforms

Comparison of soundscape metric results from simultaneously co-deployed platforms; develop tools to support this type of comparison (ie. PAMscapes). The data processing groups will establish methods for comparison, ongoing opportunistic evaluation will occur throughout the SI, and results will be added to the GitHub Soundscape Group and reported with example datasets. For SoundTrap comparisons, SoundTrap models will be specified at the time of comparison.

Support comparison across datasets with known differences

Develop a format or tool to support comparison of datasets with known differences, such as perfect timing (MARS), dropped time (SoundTrap), duty cycled (AKFSC, NWFSC), frequency-dependent calibration (NRS, HARP), file format (XWAV).