Climate Reanalyzer is very popular and deservedly so! Indeed I thought it was a large institutional project with bloated govt grants and large indifferent staff.
In terms of colors, keeping to small file size, faithful gif animation frames vs vastly smaller mp4, and importantly allowing values from the map to be read off on the palette (respectively selecting all map colors for a given value range), viewer expectations have been going up as they see examples elsewhere where someone got it all going. The wheel has been invented, let's get it rolling.
What I would recommend here is just copying the methods of L Kaleschke at AWI in producing what we've been discussing as AMSR2_AWI. Here the number of discrete colors is strictly clamped to 102 though 8-bit gif has room for 256. That is plenty. There is never a need for the 16.7 million colors of 24-bit color except with photography. Here we want binned colors. It is actually quite hard to find a set of even 20 related but distinguishable colors (for maximally unrelated, glassby).
With the AMSR2 there are no lossy jpg compressions, dithering for print, sq root two rotations of 45º to standard 'Greenland down' orientation of satellite data. The file size is merely 567 kB for 2432 x 3584 pixels. Plus it is internally geo-referenced so converts to netCDF/Panoply which people want for numeric array work. Plus viewers can easily highlight selected portions of the palette or easily replace it with the thousands of LUTs at ImageJ or even one they specify from scratch.
Three more easy suggestions, seldom seen together: use the standard 3413 EPSG WGS-84 stereographic projection for the Arctic,
specify the scale (pixel count out to special latitude of 70ºM), and provide the land mask. The isolationism we have now is an incredible waste of everyone's time; climate change above all needs data integration from disparate sources. Maps aren't pictures, they're working data.
While inclusivity (rotary dial-up, pay per MB) is laudable, it is just not the present much less the future for the 99% who might have an interest in climate reanalysis. At some point, it is not possible to cater to the very lowest devices nor assume everyone is on campus but better to find a middle ground lest the more typical users be constantly punished.
ftp://ftp.awi.de/sea_ice/product/amsr2/v103/nh/2020/10/