New Task: aniMotum/SSM processing/code, trajectory only

[MathewBiddle]

Who is requesting this?

@ioos/marine-life

What is being requested?

aniMotum/SSM processing/code, trajectory only

What is the requested deadline and why?

2024-07-01

What is the current status quo (i.e., what happens if this does not get done)?

Marine life team will not understand nor be able to reference to what and how aniMotum/SSM processing is bring done in ATN.

Provide a description or any other important information.

What indicates this is done?

Multiple actions could indicate this activity is done. The options are listed below but depend on what path the effort takes. These can be "either/or" but also "and" statements.

• [ ] A pull request to https://github.com/ioos/ioos-atn-data/ with documentation of the process.
• [ ] The source code is published on GitHub at https://github.com/ioos/ioos-atn-data/ and well documented, or other appropriate repository.

cc @mmckinzie

[mmckinzie]

What is being requested??

• (Axiom) Run all newly auto-ingested satellite telemetry data through aniMotum, use v. 1.1-04, use R-Universe Package, package will not be available on Cran, additional details available here: Fit Continuous-Time State-Space and Latent Variable Models for Quality Control of Argos Satellite (and Other) Telemetry Data and for Estimating Changes in Animal Movement • aniMotum (ianjonsen.github.io) and ianjonsen/aniMotum: fit latent variable movement models to animal tracking data (github.com)
• (Axiom) Release input and output files via the ADR
• (Axiom) Visualize output file in portal – projects pages, deployment pages, all and r/t layers (NOTE: Is it possible to make aniMotum outputs a visualization style, .i.e., hexes, tracks-raw, tracks-aniMotum?)
• (Axiom) Update portal language – update foie gras labels to aniMotum and location filter from unknown to aniMotum
• (Axiom) Work with ATN DC and code/package developer (Ian Jonsen) to troubleshoot errors/issues and improve upon base code to increase percentage of deployments which pass (currently at 70%) and improve outputted tracks, e.g., reduce angles, ensure outliers are successfully remove, including Z quality positions and positions outside of the PI reported deployment window
• (Axiom) Retroactively run all deployments through aniMotum that were previously run through foie gras, update input/output files in ADR, update visualizations in portal
• (Axiom) Publish source code and processing steps on ATN GitHub, link to ML GitHub

Deadline: First three bullets already complete (integrating into portal as a visualization style would be a new task), fourth bullet – July 1, 2024, fifth bullet – September 1, 2024+, sixth and seventh bullets – November 30,2024

[laurabrenskelle]

This issue seems related to this Marine Life Data Network Trello card entitled: "1.8.4 add provenance statement to data catalog re: Foie Gras processing to help users understand what version of the data they are accessing". Is this correct, or is this task separate and needs its own new issue? Is 1.8.4 already complete? @mmckinzie @lmshowalter3

[mmckinzie]

@laurabrenskelle This task is related to the 1.8.4 and not it is not yet complete. We used to run data through Foie Gras and now we're using the updated aniMotum code, but we still need to add provenance statements to the ATN portal to provide context to how we're applying aniMotum and why. We have updated the portal labels from Foie Gras to aniMotum but we still need to update location filter from unknown to aniMotum. We may also still need to make some minor modifications to our processing steps to ensure a higher percentage of deployments 'pass' and to improve upon some of the output visualizations. We also still need to retroactively run all developments through aniMotum, some older deployments have not been processed, while others have only been run through Foie Gras and not yet reprocessed through aniMotum.