Thank you. I make a pass over the data, find the min/max values for cooler and warmer and then use that to scale the color from full intensity (largest changes)

Thank you. I make a pass over the data, find the min/max values for cooler and warmer and then use that to scale the color from full intensity (largest changes) to least intensity (smallest change). Visually at least, it seems mostly to agree with what the clever climate scientists have made in static images. The range is calculated for the whole period so that comparing year to year should be valid. I didn't even consider blocked cookies - I set that flag in local storage since it seemed important to show some instructions the first time you see it but not thereafter. I'll experiment a bit and see if I can work around that. Edit: stars are just a bitmap wrapper around a large sphere. Cheers!