nsidc / qgreenland

Source code for generating the QGreenland package hosted at https://qgreenland.org/
https://qgreenland.readthedocs.io
Other
36 stars 9 forks source link

GrIMP ice velocity layers #749

Closed trey-stafford closed 10 months ago

trey-stafford commented 11 months ago

Description

Closes #663.

Adds annual GrIMP ice velocity mosaic layers for 2021 under the "Glaciology/Ice sheet velocity/GrIMP" group:

In total, these GrIMP layers add ~342M to the uncompressed core package. Compressed, the QGreenland core package is ~3.1G.

Styling the ice velocity vectors

I applied the style from the wind vectors layer to get started. This worked fine, but led me to realize that our wind vectors have a colormap that's inverted compared to the raster representing wind speed. This is a conflict when the two layers are visualized together. That said, the fact that the colormap is inverted makes it easier to see the vectors on top of the velocity data!

I went ahead and reversed the color map for the ice velocity vectors and forced it to match the colormap for the ice velocity magnitude raster layer. This works pretty well, but visualizing the two together is problematic because the vectors blend in with the background raster and aren't visible. Maybe an outline around the vectors would be nice, but I haven't figured out how to do that yet.

Further, I downsampled the input data for creating the wind vectors to 1.5km. This makes rendering much faster than using the full resolution data and I think it looks better too. But the downside here is that areas with major changes in velocity result in intermediate-valued vectors that, when overlaid with the raster, look like smaller/larger values than the underlying raster. Maybe a better approach would be to create vectors at full-resolution, and then 'thin' the resulting vector dataset to reduce the number of lines actually displayed?

Based on how we decide to proceed here, we should apply similar styling to the wind vectors layer.

latestScreen

Checklist

If an item on this list is done or not needed, check it with [x] or click the checkbox.

trey-stafford commented 11 months ago

I just took a look at Quantarctica, and its MEaSUREs flow vectors do not use color to represent magnitude. Just the size (thickness and length) of the vectors themselves. Maybe that's a better solution?

latestScreen

MattF-NSIDC commented 10 months ago

Just the thickness of the vectors themselves

It looks like they're using both thickness and length.