Maps revisited

I've been thinking about maps a lot lately. How can we visualize feels like weather in an intuitive way?

I've landed on using something like Uber's Hexagonal Hierarchical Spatial Index (H3) to aggregate weather data and then classify each hexagon as feeling freezing, cold, cool, perfect, warm, hot, swampy, or scorching - for the whole globe - based on your "personalized feels like" profile.

At a high level, I'm thinking:

• according to the user's current bounding box / zoom level (what you see on the map) call out to a service to individually get near real-time temp, wind speed, humidity, etc intensity grids
• aggregate these grids to intermediate output H3 cells
• apply an algorithm to the H3 intermediate output cells to calculate "personalized feels like" values for each cell, outputting the final feels like cells

In other words, raw weather grids -> transformed into aggregated hexagons -> matrix math across each weather factor (temp, wind, etc) for each location = final feels like weather hexagons.

Not sure how I'm going to do that in real-time from a technical standpoint. But that's the fun in this :D