Published April 25, 2024 | Version v1
Dataset Open

Data and Code: Detecting Blue Whale Calls in the Northeast Pacific Using Seismic Systems (Undergraduate Thesis)

  • 1. ROR icon McGill University

Contributors

Supervisor:

  • 1. ROR icon McGill University

Description

SeismoData.ipynb: This Jupyter notebook is adapted from seismosocialdistancing.ipynb created by Thomas Lecocq, Fred Massin and Claudio Satriano. SeismoData.ipynb was used to retrieve seismic waveform data from the  Incorporated Research Institutions for Seismology (IRIS) Data Management Center (DMC) (https://ds.iris.edu/ds/ nodes/dmc/) and convert files from miniSEED to SAC format. It was also used to preview waveform and spectrogram plots.

BlueWhaleDetectionResults_J53A_Dec132011.mat: This .mat file summarizes whale detection results from OBS J53A on December 13th 2011. The objective was to calibrate a detection algoritm created for Northwest Atlantic blue whale A calls by Plourde and Nedimovic (2022), so that it can target Northeast Pacific blue whale B calls using seismometers off Washington and California. Three tests were performed to find optimal parameters. BlueWhaleDetections_J53A_Test1 are the detection results of a control that uses Northwest Atlantic blue whale parameters (16.25-18Hz frequency and 68-78s period ranges). BlueWhaleDetections_J53A_Test2  are the detection results using Northeast Pacific blue whale B call parameters (14-17Hz frequency and 45-55s period ranges). BlueWhaleDetections_J53A_Test2 are the detection results using Northeast Pacific blue whale B call and C call parameters (10.5-12Hz and 14-17Hz frequency and 45-55s period ranges). BlueWhaleDetections_J53A_SCC are the 95% probability detection results from the Wilcock and Hilmo (2021) blue whale catalogue created using spectrogram cross-correlation. 

NEPBlueWhaleMATLABcodes.zip: Contains scripts to run the recurrence interval power ratio method created by Plourde and Nedimovic (2022), adjusted to detect Northeast Pacific blue whale B calls and plot waveforms/spectrograms. First run DetectBlueWhales.m to calculate the recurrence power ratio every 12 minutes, then run CreateBlueWhaleDetectionList.m to classify detections with high power ratios and likely blue whale call detections.

BlueWhaleDetectionResults_CapeMendocino_Dec15to292014.mat: This .mat file summarizes the blue whale detection results from 2 OBS (FS02D and FS07D) and 1 land seismometer (CM09A) in close proximity, using Test 2 parameters. Note if there are less than 3 BWD in a given day, these are likely false detections.

Files

NEPBlueWhaleMATLABcodes.zip

Files (182.9 kB)

Name Size Download all
md5:c8689ab43d4b74a8134b0a29615c1d5d
1.8 kB Download
md5:e955841a20e4fbcb9aadfe1bc4d5507f
6.9 kB Download
md5:750eac78e8e882a30ac6401afffdbec0
88.8 kB Preview Download
md5:80c084e304a873b419915645fb6cebf9
85.4 kB Preview Download

Additional details

References

  • Plourde, A. P., & Nedimović, M. R. (2022). Monitoring fin and blue whales in the lower St. Lawrence Seaway with onshore seismometers. Remote Sensing in Ecology and Conservation, 8(4):551–563. doi: 10.1002/rse2.261.
  • Wilcock, W. S. D., & Hilmo, R. S. (2021). A method for tracking blue whales (Balaenoptera musculus) with a widely spaced network of ocean bottom seismometers. PLOS ONE. https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0260273