High‐resolution dataset of nocturnal air temperatures in Bern, Switzerland (2007–2022).

To prepare for a hotter future, information on intra‐urban temperature distributions is crucial for cities worldwide. In recent years, different methods to compute high‐resolution temperature datasets have been developed. Such datasets commonly originate from downscaling techniques, which are applied to enhance the spatial resolution of existing data. In this study, we present an approach based on a fine‐scaled low‐cost urban temperature measurement network and a formerly developed land use regression approach. The dataset covers mean nocturnal temperatures of 16 summers (2007–2022) of a medium‐sized urban area with adapted land cover data for each year. It has a high spatial (50 m) and temporal (daily) resolution and performs well in validation (RMSEs of 0.70 and 0.69 K and mean biases of +0.41 and −0.19 K for two validation years). The dataset can be used to examine very detailed statistics in space and time, such as first heatwave per year, cumulative heat risks or inter‐annual variability. Here, we evaluate the dataset with two application cases regarding urban planning and heat risk assessment, which are of high interest for both researchers and practitioners. Due to potential biases of the low‐cost measurement devices during daytime, the dataset is currently limited to night‐time temperatures. With minor adaptions, the presented approach is transferable to cities worldwide in order to set a basis for researchers, city administrations and private stakeholders to address their heat mitigation and adaptation strategies. [ABSTRACT FROM AUTHOR]