When policymakers and city officials plan for climate change, they often rely on economic studies estimating how extreme heat affects violent crime, public health, energy systems and crop yields. But new research suggests that inaccuracies in temperature measurements could mean some of these impacts are being underestimated.
A study published in the Journal of the Association of Environmental and Resource Economists found that temperature data used in many climate-impact studies can contain substantial measurement errors. These errors can cause researchers to miscalculate the effects of temperature and, on average, potentially underestimate climate impacts by at least 15%.
The study, co-authored by Wake Forest University Assistant Professor of Economics Alex Yu and University of California San Diego Distinguished Professor of Economics Richard Carson, examined commonly used methods for estimating local temperatures. The researchers found that measurement errors can vary considerably depending on location and geography, potentially affecting conclusions about how temperature influences real-world outcomes.
Because physical weather stations are often located miles apart, researchers commonly use statistical estimates, or “proxies”, to determine temperatures in particular cities, neighbourhoods or agricultural areas. However, these estimates can differ substantially from actual local temperatures. In mountainous areas such as Boulder, Colorado, standard weather proxies differed from observed temperatures by an average of nearly 15°F, while errors in flatter, more densely monitored areas such as Chicago could be as low as 1°F.
The errors can also persist over time and vary systematically according to local geography. When estimated temperatures are linked with outcomes such as violent crime, heat-related illness or crop losses, these inaccuracies can cause the estimated effects of heat to be either exaggerated or understated. For one commonly used weather proxy, temperature measurement errors averaged 3.3°F, or 1.85°C—greater than the 1.5°C warming threshold highlighted in the Paris Agreement.
To demonstrate the potential consequences, the researchers analysed more than 20 million crime records. They first estimated the relationship between extreme heat and violent crime using city-level crime and temperature data as a benchmark. When they aggregated the same information to the county level, the estimated effect of extreme heat became smaller, showing how the geographic scale of temperature data can influence study findings.
The implications extend beyond academic research. “If climate-impact studies misestimate how severely extreme heat drives crime rates, crop loss, or health emergencies, policymakers may also misestimate the scale of the responses needed to address those impacts,” Yu said. More accurate temperature information could therefore help governments better assess climate-related risks and plan appropriate responses.
The researchers also offer practical guidance for improving future studies. No single temperature proxy performs best in every location: nearby weather stations may provide better estimates in areas with dense monitoring networks. In contrast, gridded temperature data may be more useful where stations are sparse. Expanding and maintaining weather-station networks, while more carefully matching temperature measurements with where people, farms, businesses and ecosystems are actually exposed, could reduce errors and provide a clearer picture of climate impacts.
More information: Richard Carson et al, Mismeasured and Misunderstood: Unmasking the Temperature Proxy Problem in Climate Impact Estimates, Journal of the Association of Environmental and Resource Economists. DOI: 10.1086/742106
Journal information: Journal of the Association of Environmental and Resource Economists Provided by Wake Forest University