Week 06 · September 28–30
Follow the smoke across the border.
Launch the wildfire smoke project by tracing each link from a Canadian fire to a U.S. county, then finish Classwork 4 by transforming, joining, and validating data before interpreting any pattern.
This week
Wildfire smoke
- Explain why smoke from Canadian fires is an unpriced cross-border externality.
- Explain why a simple smoke–death correlation can be too large or too small.
- Describe how distant fires carried by the wind create a natural experiment.
Measurement
- Keep units visible: µg/m³ of smoke, deaths per 100,000, and person-days.
- Align time carefully: UTC days versus local days, daily versus monthly data.
- Treat suppressed counts and disagreeing smoke products as measurement problems.
Transform in R
- Reshape, summarize, and join data with explicit keys and row grains.
- Validate row counts, join relationships, units, and denominators.
- Describe each pattern within the sample, year, and scope of the data.
Lecture slides
Research guide
When Canadian smoke crosses the border
Read the guide before choosing a contribution path. It maps each link from fire to value onto the public data that can measure it, and it includes the evidence-memo template.
Classwork
Classwork 4: Governance and Climate Finance in the Developing World
Finish the core path—Parts 0, 2A, 3, 4, and 7—then extend it with the distribution, ranking, uncertainty, six-dimension, and map displays in Parts 5–6 and 8–10. Keep the sample, year, unit, join key, and data scope visible in every table and figure.
By the end of Week 6
Prepare the following:
- one contribution path chosen from the wildfire research guide;
- one smoke episode for your evidence memo, with its dates, source fires, and affected area;
- the completed Classwork 4 figures, each stating its sample, year, unit, and data scope; and
- one unit, timing, or missing-data issue that must be resolved before comparing places.