Do large-area-average temperature series have an urban warming bias?

Conclusions

The arguments presented by Wood and his criticisms of the methods used by Jones et al. are largely fallacious and are generally based on misconceptions and unwarranted assumptions. This does not, of course, mean that the Jones et al. data are perfect. Jones et al. (1986a, b) have clearly stated that, at the regional level, there may be residual uncertainties in their gridded data set. Furthermore, they have noted that there is a residual uncertainty in the global-mean change since late last century of ±0.2 °C (Wigley et al., 1986), although the main reasons for this uncertainty do not relate to urban warming. Further studies of possible urban warming biases are certainly warranted, particularly at the regional scale. However, for averages over continental-scale areas upwards, it is unlikely that any significant urban warming bias remains.