Coexisting with Whales

Interactive map of cetacean observations and vessel strike risk in Norwegian and surrounding waters. A pre stage screening tool for understanding where whales and ships overlap, and how cutting vessel speed changes strike lethality.

Live demo: https://coexisting-with-whales.no/

The GitHub Pages URL https://almazermilov.github.io/coexisting-with-whales/ redirects to the canonical domain above.

A sister project to Coexisting with Birds, inspired in part by the HUB Ocean initiative and the Mesh Oslo "what else can you build" prompt.

Quick start

python3 -m http.server 8081
# open http://localhost:8081

No build step. Static files served from any HTTP server. The site loads Leaflet, Leaflet.heat and CARTO basemaps from CDNs.

Running tests

npm install
npx playwright install chromium
npm test                 # vitest unit + playwright e2e
npm run test:unit
npm run test:e2e

Features

• Heatmap view: cetacean observation density across the Norwegian Sea, Barents Sea, North Sea and Greenland Sea using a perceptually uniform ocean palette.
• Points view: individual observations colored by species; click a point for details (red list status, common names, source dataset).
• Species filter: 25 cetacean species, with quick filters for great whales, threatened species, red-list species and high strike risk species.
• Month slider: see seasonal variation including the summer feeding aggregations (June to September) and the winter herring fjord aggregations (November to January).
• Major ports: 25 Norwegian ports rendered with risk-coloured markers (icon size scales with annual cargo throughput).
• Sea regions: click a sea region to open a per region calculator with a vessel speed slider that updates strike lethality live.
• Per port collision calculator: click any port for a Vanderlaan and Taggart 2007 lethality curve, the list of vulnerable species nearby and the moderated risk score.
• EMODnet vessel density overlay: 1 km resolution shipping density grid from EMODnet Human Activities, togglable via a single checkbox (no API key, no auth).
• Sea region density layer: each region coloured by observation count to show data confidence.
• Literature popups: every formula and dataset has a hover popup with the citation and a DOI hyperlink (Vanderlaan and Taggart 2007, Conn and Silber 2013, Rockwood et al. 2017, Williams and O'Hara 2010, Redfern et al. 2013, Laist et al. 2001, Nisi et al. 2024 and more).
• Light and dark themes: switch between ocean night mode and a daylight chart style; the map basemap changes with the interface theme.
• Hide UI: toggle all overlays with a button or the [H] key for a clean map view.

Datasets

Cetacean observations

• Source: GBIF (Global Biodiversity Information Facility)
• API: https://api.gbif.org/v1/occurrence/search
• Scope: order Cetacea (taxonKey 733), bounding box lat 56-82, lon -10 to 70
• Sample: 10,000 records out of approximately 138,000 in the bounding box
• Filter: excludes MACHINE_OBSERVATION (acoustic hydrophone clusters that distort the heatmap)
• License: CC BY 4.0, CC0 1.0 or CC BY-NC 4.0 (per contributing dataset)
• Citation: GBIF.org GBIF Occurrence Search for order Cetacea, Norwegian and surrounding waters

Sea regions

• Source: Marine Regions IHO Sea Areas v3 (DOI 10.14284/323) via VLIZ WFS
• Scope: Norwegian Sea, Barents Sea, North Sea, Greenland Sea
• License: CC BY 4.0

Norwegian EEZ

• Source: Marine Regions Maritime Boundaries Geodatabase v12 (DOI 10.14284/632)
• Scope: Norway mainland (MRGID 5686), Jan Mayen (8437), Svalbard (33181)
• License: CC BY 4.0

Coastline

• Source: Natural Earth 50m physical vectors
• License: public domain

Major ports

• Source: hand curated from Wikipedia "List of ports in Norway" and Kystverket
• Coverage: 25 ports including Oslo, Bergen, Stavanger, Trondheim, Tromso, Bodo, Hammerfest, Kirkenes, Mongstad, Karsto, Sture and others
• Fields: name, latitude, longitude, type, approximate annual throughput in million tonnes

Vessel density (live overlay)

• Source: EMODnet Human Activities Vessel Density Maps
• API: WMS at https://ows.emodnet-humanactivities.eu/wms, layer vesseldensity_allavg
• Scope: 1 km grid covering all EU waters and neighbouring areas (2017 to 2024 annual average, all vessel types)
• License: EMODnet open with attribution to CLS, ORBCOMM and vesseltracker.com as AIS data providers

Conservation status

• Norwegian Red List 2021: Artsdatabanken, with verified entries for beluga (EN), bowhead (EN), narwhal (VU), blue whale (EN), fin whale (LC nationally despite VU globally), harbour porpoise (LC), killer whale (LC), sperm whale (NA, since reproduction does not occur in Norwegian waters)
• Global IUCN Red List: IUCN, used as a fallback when the Norwegian assessment is NA
• OSPAR list of threatened and declining species: bowhead, North Atlantic right whale, blue whale, harbour porpoise

Collision risk literature

The screening score uses peer reviewed formulas. The full bibliography is rendered inline in the info modal with DOI hyperlinks. Headline references:

• Vanderlaan and Taggart (2007). Vessel collisions with whales: the probability of lethal injury based on vessel speed. Marine Mammal Science 23(1), 144 to 156. doi:10.1111/j.1748-7692.2006.00098.x
• Conn and Silber (2013). Vessel speed restrictions reduce risk of collision related mortality for North Atlantic right whales. Ecosphere 4(4), art. 43. doi:10.1890/ES13-00004.1
• Rockwood, Calambokidis and Jahncke (2017). High mortality of blue, humpback and fin whales from modeling of vessel collisions on the U.S. West Coast. PLOS ONE 12(8), e0183052. doi:10.1371/journal.pone.0183052
• Williams and O'Hara (2010). Modelling ship strike risk to fin, humpback and killer whales in British Columbia. J. Cetacean Res. Manage. 11(1), 1 to 8.
• Redfern et al. (2013). Assessing the risk of ships striking large whales in marine spatial planning. Conservation Biology 27(2), 292 to 302. doi:10.1111/cobi.12029
• Laist et al. (2001). Collisions between ships and whales. Marine Mammal Science 17(1), 35 to 75.
• Nisi et al. (2024). Ship collision risk threatens whales across the world's oceans. Science 385, 1136 to 1141. doi:10.1126/science.adp1950

Refreshing data

python3 fetch_data.py             # GBIF cetacean observations (~1 min)
python3 scripts/fetch_geo.py      # Sea regions, EEZ, coastline, ports (~30 sec)

fetch_data.py stratifies across months to balance the strong summer skew. Adjust limit_total to change the sample size; 10K records is roughly 1.7 MB. The hard pagination ceiling on the GBIF search API is 100K offset; for larger samples use the asynchronous Occurrence Download API.

scripts/fetch_geo.py requires mapshaper (npm install -g mapshaper) for simplification. Without it the geographic files will be much larger.

File structure

index.html                Main HTML with metadata, side panel, modals, legend
css/style.css             Styles with CSS custom properties (ocean palette)
js/data.js                Constants: species, IUCN, Norwegian Red List 2021, seasonality
js/refs.js                Literature references with DOI links and tooltip helpers
js/scoring.js             Pure functions: peer reviewed collision risk physics
js/ui.js                  DOM updates: filters, species list, toggles
js/app.js                 Entry point: map init, data loading, events
data/whales_norway.json   10K cetacean observations from GBIF
data/sea_regions.geojson  Norwegian, Barents, North, Greenland Sea (IHO v3)
data/coastline.geojson    Natural Earth 50m, clipped to Norway bbox
data/eez.geojson          Norwegian EEZ (mainland + Jan Mayen + Svalbard)
data/ports.json           25 major Norwegian ports (hand curated)
fetch_data.py             GBIF cetacean fetcher
scripts/fetch_geo.py      Geographic data fetcher with mapshaper simplification
tests/unit/*.test.js      Vitest unit tests (64 tests, scoring + filters)
tests/e2e/app.spec.js     Playwright end to end tests

Limitations

This is a screening tool, not an environmental impact assessment.

• Cetacean records reflect observer effort, not true density. Coastal viewpoints (Andoya, Lofoten, Tromso) and tourist boat routes are overrepresented.
• EMODnet's vessel density grid is a 2017 to 2024 annual average, not real time.
• Norway specific peer reviewed strike risk modelling is sparse. We apply global lethality physics (Vanderlaan and Taggart 2007) to a global set of species accounts. For real planning use site specific surveys, Kystverket / BarentsWatch live AIS, and modelled species distributions from Havforskningsinstituttet.
• The North Atlantic right whale (CR) is included for completeness but is functionally absent from the NE Atlantic.

Deployment

Hosted on GitHub Pages, free. Deployment is automated via .github/workflows/deploy.yml on every push to main. Tests run via .github/workflows/test.yml. No external dependencies (no Vercel, no AWS, no API keys).

License

MIT, see LICENSE. All data is from open sources with attribution preserved.

Author

Built by Almaz Ermilov in Oslo, April 2026.