Short snippet of full smoke animation from article
Watch July's wildfire smoke travel across the country (breaking news)
For this story, I scraped the smoke data and collaborated with my colleague on collating the scraped images into an annotated animation (see below for more about this technical challenge). I also conducted the interviews and wrote the accompanying article to explain the spread you see in the video.
This story was designed to help readers understand the far reaching effects of explosive wildfires that are more likely in a warming world.
This story took 2 days to complete.
Tools
Python (wrote scraping script to run NOAA smoke model and generate 700 map PNGs)
Adobe Premiere Pro
d3.js
The scraping script at work (Nobody is moving the mouse or typing!)
The scraping script had to move the mouse like a human does because the NOAA website does not generate the map corresponding to that day/hour’s model run unless someone directly clicks on it and sends the request (ie. the map images are dynamically generated). To have hourly frames that could be appended into an animation this meant we needed to click on over 700 links! So I wrote a script that took over my mouse and keyboard overnight and scraped everything we needed. It had to have lots of error catching for when spotty internet or a buggy server got in the way.
