The starting question was narrow: can a heatmap animation on a map hold up as a serious visualization, rather than just a coloured picture? Temperature is a continuous field with no borders, so neither a choropleth nor a point map fits it well. This chart was an attempt to render that continuous spread — where the heat concentrates and how the hot zone shifts across the period — on a single readable surface.
What the animation makes visible is reach rather than record. The author's reading is that the interesting part is spatial: a broad band of heat forming over the west and drifting east across the continent, rather than any single peak figure at one location.
Values are modeled reanalysis output on a 1° grid, not observations from weather stations. A single grid cell stands in for roughly 100 km of terrain, so mountains, coastlines, and local variation are smoothed away. The grid was deliberately coarsened this far to keep the animation running in a browser; finer resolution stalled it.
A heatmap renders colour from the combined density of surrounding points rather than from raw values. Colour intensity therefore reflects how tightly hot cells cluster, not a fixed temperature at any one spot, and it should not be read as a temperature scale.
This is an absolute-temperature map, not an anomaly map. It shows daily maximum temperature as it was, not how far above or below normal any day ran. Nothing here is computed against a long-term baseline, so the chart cannot answer "how much hotter than usual" — only "how hot, and where."
The colours are not a readout of exact degrees, and at a 1° grid no city-level temperature can be taken from the map; that would require each country's meteorological observations. The chart also says nothing about causes. Long-term warming is understood to raise heatwave risk in general, but attributing this specific event to climate change would require separate attribution study, and any national record temperatures fall outside this dataset.