Very interesting, thanks Jim.
It took me a while to reconcile the coarse & fine scales for both temperature & salinity profiles because they distort the intermediate values for each colour in going between coarse & fine, while keeping both endpoints the same.
For the example of temperature, the endpoints are violet at -2.0oC and red at +1.8oC, applying for both the coarse and fine scales. However the fine/coarse values respectively for 0%, 20%, 40%, 60% and 80%, 100% up the colour scale are, in degrees C, -2.0 for both (0%, violet), -1.6 vs 0.0 (20%, dark blue), -1.2 vs 0.4 (40%, light blue), -0.8 vs 0.6 (60%, light green), -0.4 vs 0.8 (80%, yellow), 1.8 for both (100%, red).
Why do they do that!? Presumably to give slightly better resolution for both plots but... ugh!