I'm working on a visualisation of weather, including wave heights. The source data is from a WaveWatch 3 model, and I'm using GeoServer 2.7 (with NetCDF and dynamic colormap extensions) to do the rendering. My question relates to how to best represent that data to users who have domain expertise (e.g. commercial shipping / deep water fishing / Navy) but not much numerical forecasting expertise.

The use of the data isn't (completely) known, and will vary by user. One example is that training for helicopters or aircraft might not be allowed above seas that are above some (specified) height. Alternatively, it might be some relative indication with a visualisation over time (e.g. so fishing can occur in the smoother water to the west, before moving to the east after that area settles).

The end result will be WMS, which will be rendered above the basemap (with opacity that is user selectable from 0 to 100%). The basemap is user selectable on the client side and I have no way to know what the basemap colour will be (best guess: some solid colour for deep ocean, perhaps blue, white or black).

There is a good question (and some excellent answers) on an appropriate colour-ramp for DEM, which was informative. However most of the answers for that were related to making the terrain "look right" (e.g. with hill shading), which isn't really appropriate for something like wave height.

My priority is to communicate the weather conditions first, and to look good (as a close) second.

My question is: what is an appropriate wave height colour ramp? If there was an accepted international standard (or some fairly common national standard) I would likely prefer that. Failing that, I'd go with accepted designs for this kind of visualisation.

  • Really interesting question. What would wave height communicate to the viewer? Do you need to show intervals ('calm', 'choppy', 'terrifying') or a continuous scale? – Simbamangu Jan 10 '15 at 6:33
  • I added some examples. In this case, both sorts of styles might be interesting to different users / viewers. – BradHards Jan 10 '15 at 8:07
  • @Simbamangu Wave height is very important for safety, but the combination of height, period etc to give the terrifying aka Phenomenal en.wikipedia.org/wiki/Sea_state – MichaelStoner Jan 15 '15 at 9:00

I'd go with the colour ramp from NOAA, from the suppliers of your data,

Example for Significant Wave http://polar.ncep.noaa.gov/waves/viewer.shtml?-multi_1-NE_atlantic-

I've spoken to a supplier of weather assets about this and their comment was "Pick your own colours" there's no standard.

There is some interpolation going on in the NOAA viewer, and your going to be using bicubic interpolation within geoserver? Using the same ramp as NOAA will allow very quick confirmation that the whole processing chain is working correctly, and when this software goes into maintenance mode it's great that the supporters have an external render of the same data.

If it's to be shown over different base maps, then the style/colour ramp can change too. You create styles dependant on the base map, so you create more color ramps and styles called 'OverlaidOnBlackBaseMap', 'OverlaidOnBlueBaseMap', 'OverlaidOnWhiteBaseMap', and when the basemap changes the style passed to the overlay changes too.

Comment: You've mentioned dynamic colormap, which I've not investigated, but I take it you're not going to have a dynamic colormap? If the ramp were to go green 0 meters to Red being 15m high, but on a calmer day the highest value in the data was 5m you'd not want that raster layer to be dynamic and render the 5m as red.

  • Thanks for the suggestions. Dynamic colourmap can be based on a fixed extent - its usually done based on gdal summary of the data, but it doesn't have to. The case for dynamic colourmap is to show "better and worse" areas rather than any objective measure of height. The complexity is that its usually for all of the data, not the part of the data that is rendered on the map. – BradHards Jan 15 '15 at 2:22
  • Ah, thanks for the reply wrt to the dynamic colourmap. I read in too much hast your original question where you explained that it's not an absolute style aka "standard" weather forecast map, but rather a visualisation highlighting the "best" area for a particular "task". – MichaelStoner Jan 15 '15 at 8:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.