The debate over whether urbanization and related socioeconomic developments affect large-scale surface climate trends is stalemated with incommensurable arguments. Each side can appeal to supporting statistical evidence based on data sets that do not overlap, yielding inferences that merely conflict with but do not refute one another. I argue that such debates can only be resolved in an encompassing framework, in which both types of results can be demonstrated on the same data set, in such a way that apparent support for one conclusion occurs as a restricted case of a more general specification that supports the other, and where the restrictions can be tested. The issues under debate make such data sets challenging to construct, but I give two illustrative examples. First, insignificant differences in warming trends in urban temperature data between windy and calm conditions are shown in a restricted model whose general form shows temperature data to be strongly affected by local population growth. Second, an apparent equivalence between trends in a data set stratified by a static measure of urbanization is shown to be a restricted finding in a model whose general form indicates significant influence of local socioeconomic development on temperatures.

You don’t say.

I should add that every time I see a video of earth from space like this, I just shake my head in amazement at the people who think that there’s no market for people to go into space.