Color blind assessment
We should perform an assessment of all the charts that we generate and evaluate how a color-blind person would see them. ~10% of the world population is color-blind. If some of the charts becomes not interpretable, then we should change the color scheme.
Seems like a great idea, but how do you perform such a test? Do you have any suggestions? Are there online tools for this?
—
Hi, I have found this website that can simulate various types of color blindness and I have run it against a composite of all the legends. results are in the attached PDF. I am not 100% sure how to evaluate te results, most seem fine too me, the most problematic seems to be the monochomatic, especially for the UTCI scale, where I am using the same colors as in the original paper.
Hi, Another test which may enhance the pdf generated. Microsoft Edge has a built in vision deficiency emulator which can alter browser output based of a variety of vision/colour blind impairments. https://learn.microsoft.com/en-us/microsoft-edge/devtools-guide-chromium/accessibility/test-color-blindness
I am happy to test the emulator against all the charts and generate a report if deemed worthwhile @FedericoTartarini
Issue110_Color_Blind_assessment.pdf
Attached are the vision impairment emulator results for all graphs in Clima. The attached report shows output for the following vision impairments conditions:
- Reduced contrast
- Protanopia (no red)
- Deuteranopia (no green)
- Tritanopia (no blue)
- Achromatopsia (no colour)
The charts that have some issues and may need some changes are:
- [ ] heatmap with Achromatopsia (no colour) since both sides of the colourmap appear to be bright
- [ ] solar cloud coverage with Achromatopsia (no colour) both two colours look the same
- [ ] wind tab second hatmap Deuteranopia (no green), Protanopia (no red), and Achromatopsia (no colour)
- [ ] outdoor comfort heatmap Achromatopsia (no colour)
- [ ] wind rose graphs with Protanopia (no red). Some of the blues, look kind of the same