Reading a Temperature Record with Confounders
A regional weather station has recorded daily maximum temperatures since 1950. The raw record shows a warming of 0.9 °C over 70 years. However, the station was moved 4 km from a rural site to an airport in 1978, and the surrounding town grew from 8,000 to 55,000 people over the same period. A nearby rural station, unaffected by urban growth, shows 0.5 °C warming over the same period with a similar instrument history. A student concludes the 0.9 °C trend proves regional climate change is accelerating.
1. Evaluate the student's conclusion using the two records, and explain the role of the station move and urban growth.
Read the worked answer
The student's conclusion is not supported as stated.
Confounders:
- The 1978 move to an airport introduces a site change; airport surfaces (tarmac, buildings) absorb and re-radiate heat, creating a local warm bias.
- Urban growth from 8,000 to 55,000 intensifies the urban heat island, which raises measured temperatures independent of regional climate.
Comparison: the rural station shows 0.5 °C, the town station 0.9 °C. The 0.4 °C difference is plausibly explained by urban/site effects rather than regional warming. The rural record is the better proxy for regional climate.
On 'accelerating': a single linear trend over 70 years cannot demonstrate acceleration. Acceleration requires comparing trends across sub-periods (e.g., 1950–1985 vs 1985–2020) and showing a statistically significant increase, ideally with homogenised data. The student has conflated a local bias with a regional signal and a single slope with acceleration.
2. Describe how the temperature record should be adjusted and what evidence would be needed to support a claim of regional warming.
Read the worked answer
Adjustment (homogenisation):
- Identify breakpoints such as the 1978 move using station metadata and statistical change-point tests.
- Compare the town station with neighbouring rural stations to estimate the urban/site bias, then subtract it.
- Document each adjustment so the record is reproducible and auditable.
Evidence for regional warming:
- A consistent trend across multiple independent stations, including rural and coastal sites.
- Agreement with other variables: minimum temperatures, frost days, growing-season length, sea-surface temperature.
- Statistical significance and confidence intervals on the trend.
- Physical consistency with known drivers (greenhouse gas forcing), not just one station's slope.
Key point: a defensible claim rests on spatial coherence and multiple lines of evidence, not on a single unadjusted urban record.