Paper Reviewed
Levermore, G., Parkinson, J., Lee, K., Laycock, P. and Lindley, S. 2018. The increasing trend of the urban heat island intensity. Urban Climate 24: 360-368.

Every now and then we come across a new journal article that reminds us of the challenges that exist in finding an anthropogenic signal within the near-surface temperature record. The latest study to cross our desk that gives us pause in this regard is from Levermore et al. (2018).

The essence of their work included an examination of the urban heat island in Manchester, UK. Using hourly temperature data from an urban and rural (~12 km away) UK Met Office site over the period 1996-2011, the five scientists calculated a statistically significant urban heat island influence of 0.021°C per year in the Manchester data, which trend is "approximately equal to the lower predictions of climate change." Levermore et al. also calculated the change in urban morphology between 2000 and 2009 based on changing aerial green space, determining that "the green area has reduced by up to 11% over the whole area shown although it is only a 1.5% reduction within 200 m of the [urban] station." And it was this reduction in green space that was ascribed by the authors to be the main factor of urban heat island contamination in the Manchester temperature record.

Think about the above results. A 1.5 to 11% reduction in green space over a ten-year period was significant enough to raise temperatures in the Manchester record by an average of 0.42°C. That spurious warming is about half the magnitude of warming that the world has seen globally since the end of the Little Ice Age over a century ago. Clearly, correcting for urbanization effects is a knotty issue of climate science and, if not done correctly, will add spurious warming unrelated to natural or anthropogenic trends. And that is one of the reasons that satellite-derived temperature data are considered to be superior to data collected at near-surface land-based locations.