orca-dclde

Open data related to the endangered Southern Resident Killer Whales (SRKW) to support a Detection, Classification, Localization, and Density Estimation (DCLDE) workshop.

Join our Discussions and check out our Wiki! As of summer 2023, the Wiki covers:

• Competition Evaluation
• Data Sharing Agreement
• To Do
• Training Data

Initial challenge ideas (archived open data opportunities)

General

1. Holger thought (via Kaitlin): expand to global killer whale ecotypes?
   1. Holger suggestion of Norwegian KW annotations?
   2. Scott Q: did anyone ever annotate KW signals in the PALOA archive from Antarctica?
   3. Scott: Ask re Icelandic recordings/annotations?
   4. Scott: Reconnect with Dan Olsen et al. re Gulf of Alaska ecotypes?
   5. Reach out to Russian KW bioacousticians?
2. Make stronger connections with past WA/BC and compared annotated data sets with past published work? Are we missing any opportunities?
   1. NRKWs: Miller's localization work? Volker Deecke's work? Christian Bergler's efforts with NRKW calls from OrcaLab?
   2. SRKWs: Andy Foote's SRKW work? Monika Wieland's undergrad thesis and subsequent work? Candice Emmons? David Bain?
   3. Bigg's: How many annotations or call type examples does John Ford have?

Detection

Supervised

1. A subset of Orcasound archives (~2018-present, ~3-5 locations in WA, 48kHz, lossy HLS data) and Orcasound bioacoustic bout log
2. SMRU archives (Lime Kiln, WA...) and associated bioacoustic event log (from Jeanne Hyde+?)
3. SIMRES archives and associated field observations at East Point, Saturna Island, BC

Unsupervised

1. Oregon coast OOI archives (winter presence data gap)
2. Neptune ONC Folger deep archives
3. Venus ONC Strait of Georgia archives

Un/quasi/supervised?

1. MANY autonomous deployments in BC habitat of SRKWs, especially by DFO (Thomas, James, John Ford; Shiela? Harald?)
2. MANY autonomous deployments in WA habtiat of SRKWs, especially by NOAA (Marla, Brad; validated by Candi)

Classification

1. Orcasound open training data
2. HALLO open training data (URL TBD, fall 2022)
3. Online SRKW call catalogue (v1 | v2 | v3, URL TBD fall, 2022)
4. Online SRKW whistle catalog? (Partner with Monika of OBI?)
5. Online SRKW echolocation catalog?
6. Standardization of annotation scheme for SRKW signals (Draft Transboundary Dictionary)
7. Should percussives be included as a SRKW signal?

Localization

1. SMRU CAB data? (Were any collected simultaneous to a DTAG deployment to validate 3D localization?)
2. ULS data (tetrahedral arrays) from Strait of Georgia (ONC) and/or Boundary Pass (ECHO/JASCO?) and/or ??
3. Long-baseline opportunities? (SIMRES East Point vs Monarch Head? Orcasound Lab vs Lime Kiln, e.g. Bigg's kill observations by CWR

Density Estimation

1. Cool thing about SRKWs: we know the total population perfectly.
2. Increasingly, we have visual observations that can constrain the number of individuals within detection range
3. DE at pod level could be done in many places (maybe best where localization could also offer constraints?)
4. DE at matriline level probably needs visual presence and ID info from a dense network observing continuously during daylight hours (e.g. west side San Juan Island [Facebook group, Lime Kiln staff, Oceans Initiative theodolite team; or maybe East Point?)

Overarching questions:

• What DTAG audio recordings are available (from Holt in US or ?? in Canada) and how might they best be used?
• Where have/are SoundTraps or other recorders being deployed (in e.g. 2022 in WA and/or BC) and how could they be leveraged? (SanctSound deployments in Puget Sound?)