| If you are one of our readers based in the US, chances are you have experienced sweltering heat over the past few days. And depending on where you are, this hot weather may well persist into next week. Nearly 100 million in America are living in areas that have experienced extreme temperatures over recent days, with major cities such as Boston and New York declaring heat emergencies and cancelling public events. The heat dome causing these highs is due to move southwards, bringing record temperatures to the mid-Atlantic and southern states, according to NOAA’s National Weather Service. But this isn’t a localised event. Extreme weather is being experienced across the globe. In Saudi Arabia’s Mecca, hajj pilgrims have reportedly died from the hazardous levels of heat. Further east, India is continuing to bake in prolonged heatwaves that began as early as April. Mexico has been hit by torrential rain brought by tropical storm Alberto after an extreme heatwave in May that thrust 90% of the country into drought conditions. The prolonged heat had been made 35 times more likely by climate change, said the World Weather Attribution academic research group in their latest scientific study published this week. How we made it We’ve produced countless temperature and anomaly maps, so I wanted to take a slightly different angle on what is becoming a familiar story by visualising the number of people now being affected by the US heatwave. So for this week’s climate graphic, I introduced population figures to our typical climate data and tested out how we might show this relationship. In an early experiment, I produced a population spike map and overlaid it the daily maximum temperatures. Colours represent daily highs, and spike height shows population. Copernicus’ Global Human Settlement Layer is a useful resource for high-resolution, gridded population data — you can download high-resolution tiles and they provide population forecasts for 2025 and 2030. Nasa’s GMAO provided the hourly surface temperature data, from which I calculated the daily maximums (we’ve talked about working with this data in a previous newsletter, if you email us at climate@ft.com we can forward you this edition). Using Blender and the Qgis software, I produced the draft spike map, but after some quick tests abandoned this route. The render below is very much unfinished, but even with better camera angles and improved lighting, the population spikes seemed to interfere with the temperature data too much, making both variables difficult to read, rather than elegantly overlaying one over the other. Instead, I used population data aggregated at the county level and pulled this along with the daily maximum raster data into Qgis. The ‘zonal statistics’ function, allows you to calculate relevant stats from raster data for overlaid polygons (in my case a US county shapefile). This is how I obtained the average daily highs at the county level. Finally, after calculating the centroid of each county, I used a scaled circle approach to show population size and a sequential colour palette for the corresponding maximum temperatures. To draw the viewers’ focus some more, I chose only to include counties where the maximum had exceeded 26C on a given day. In a final test, I created a gif that cycled through the daily data, but in the end decided to create a small multiples map, which allows viewers to interrogate the maps in more detail, with added annotations and labels to explore. In climate data visualisation, it is often a challenge to incorporate the human impact of extreme weather events. Visualising the number of people affected is a small step in attempting to do so, but I’m sure there are many more creative techniques and approaches to be explored. As northern-hemisphere summer is sure to bring more heatwaves over the next couple of months — perhaps along with other extreme weather events — this is a challenge to keep in mind. If you have recently come across any good examples of climate data visualisation highlighting the human impact, please do share them with us (climate@ft.com) and we may feature them in our next newsletter! Three of the best | 
