Each city is compared against itself, and that is the design detail worth noting. Within every district and independent city, sub-areas were ranked by each sociodemographic factor, and temperatures at the top and the bottom of that ranking were compared [7]. Urban heat island intensity was worked out separately, by holding city temperatures against neighbouring rural areas with similar topography [5]. Those two choices remove most of what would otherwise manufacture a correlation, regional climate and city size included, because the comparison never leaves the city it starts in.
The summary omits a denominator. The work covers all German districts and independent cities [4], and KIT reports the night-time difference in 84.2% of the assessed cities [1], but the number of cities behind that percentage is not stated [19]. Inverted, 15.8% of them, about one in six, did not show it [1]. Nor is there a magnitude: the report says only that night air was significantly warmer, without giving a figure in degrees [21].
The temperature record spans three summers, and the census is a single 2022 snapshot sitting in the middle of them [4][6][2]. A reasonable pairing, though it treats neighbourhood composition as fixed across a window in which people move.
Both authors are firm about the limits. Benz says that without high-resolution income data the team cannot tell whether the differences trace to socioeconomic inequality [10]. The physical candidate the group names is ordinary enough: high building density and few green spaces, in a fabric whose sealed surfaces store heat by day and radiate it at night [12][17].
Then there is the measurement question, which is where this reaches procurement. Land surface temperature is what satellites see well, and it describes how hot roofs and asphalt get. The two-metre air temperature is what Chawla says contributes more directly to the heat stress people are exposed to, and is the measure most shaped at night by buildings, green spaces and materials [14]. The two do not always reveal social inequalities in the same places [13]. A city that maps its heat from surface data and a city that maps it from night air will not always end up treating the same blocks.
The data do not show whether anyone was harmed. The output is temperature, and the material carries no health outcomes and no indoor measurements [20]. Exposure is the first term in that chain rather than the whole of it.
One reassuring sign sits in the results themselves: the three factors do not line up in the same direction [1][8][9]. Had they, the likeliest reading would be that the method was rediscovering dense inner-city districts under three different names.
Copernicus recorded June and July 2026 as warmer than any other two-month period on record [16], and Chawla notes the disparities are sharper in cities than in suburban or rural areas [15]. For a council with a fixed budget for trees and shading, a correlation that survives within-city comparison is a usable ranking device, which is the most this evidence will bear and more than most municipalities currently have to work with.
