I wanted to create a visualization of the scale of the ongoing wildfires in New Mexico, but could not find a source of timeseries data for them.

This library scrapes the HTML of the Incident Information System and applies (rough) NLP in order to extract the area and containment of a fire given its incident ID.

It also supports invocation through the command line to plot the area and containment of a single fire, or dump that fire's data to CSV.

from inciweb_timeseries_scraper import get_all_data

data = get_all_data(8103)
# data now contains (date, area, containment) tuples

The ID 8103 comes from the URL of the incident page, e.g. https://inciweb.nwcg.gov/incident/8103/.

# display a plot of the area and containment time series
$ python3 -m inciweb_timeseries_scraper plot 8103

# plot to file
$ python3 -m inciweb_timeseries_scraper plot 8103 -o black_fire.png

# save data to csv
$ python3 -m inciweb_timeseries_scraper csv 8103 -o black_fire.csv

• The data posted on InciWeb comes from different sources and the NLP techniques I'm using rely on some heuristics that might not work for all events.
• This will break immediately if the HTML of InciWeb changes.
• Making many automated requests like this may eventually cause you to get blocked or throttled.
• Some posts that are not actually daily updates may have spurious data extracted from them and thus will need to be manually removed.