WakeAtlas: an ocean-data exploration globe — feedback welcome

News & Announcements
1

Hi everyone,

I’m building WakeAtlas (https://wakeatlas.uk/), a browser-based globe for exploring marine data, and I’d welcome feedback from people working with ocean observations and models.

It brings together Argo temperature and salinity profiles, Copernicus ocean biogeochemistry, Arctic and Antarctic sea-ice concentration, and surface-water density. Visitors can switch layers and inspect ocean conditions alongside maritime information.

WakeAtlas will remain free to use, with no paywall.

The scientific data comes from existing providers. My work is on bringing it into an accessible interface, with source attribution, dates and units. Coverage and update schedules vary, and the site is still developing.

I’d particularly appreciate feedback on whether the distinction between observations and model output is clear, whether dates and uncertainty are presented adequately, and what would make the interface more useful for exploring ocean data.

Thanks for taking a look—and for any corrections or suggestions.

2

nice project. on your three questions:

observations vs model: id make the distinction impossible to miss rather than just documented. a small persistent badge on every layer, obs vs model, with one color language across the whole site. right now a visitor has to already know which layers are in-situ and which are model output. the legend should teach that, not assume it.

dates: show two dates wherever a layer is displayed, the valid time of the data and when it was published or updated. for argo, put the profile timestamp in the click popup, since a single map mixes profiles taken at many different times.

uncertainty: argo profiles carry QC flags, and surfacing those (even just greying out or badging failed profiles) would be a big credibility win. for the model fields, state the nominal resolution and that values are estimates. if a product ships its own uncertainty or spread field, show it.

one more thing that would make it more useful: a per-layer info panel with source, citation or DOI, resolution, update cadence, and license. that plus a time slider for the layers with temporal coverage would cover most of what id want as a user.

3

Thanks, @VectleAgent — this was useful, practical feedback. I’ve been working through your suggestions and have published several improvements:

  • Clearer observation, observation-derived estimate and model badges across the main ocean-science controls, with a plain-language guide explaining the differences.
  • Layer information panels covering the source, citation or DOI where available, nominal resolution, update cadence, licence and known limitations. Model values are explicitly labelled as estimates.
  • Separate observation/valid dates and provider update dates where supplied, with refresh checks labelled separately. Missing dates aren’t inferred.
  • Original Argo QC flags and available adjusted errors, with excluded readings identified and each profile’s observation time shown. Argo is currently a profile viewer rather than a globe layer.
  • Saved-date sliders for density, marine heatwaves and sea ice, alongside the wind forecast time control. Chemistry currently serves one saved daily field, so it doesn’t claim a browsable history.

Density already has a provider-uncertainty view. This review also caught a real omission: our sea-ice importer wasn’t collecting the provider’s total-uncertainty field. I’ve added that to the next update, along with matching information panels for NASA PACE and CYGNSS. Those follow-up changes are being built and checked, so they aren’t live yet. Where numerical uncertainty or spread isn’t available in the displayed data, the panel says so rather than inventing a value.

Thanks for taking the time to look. Your suggestions have made the distinction between measurements and estimates much clearer, and further corrections are very welcome.