Improving Urban Temperature Measurements and Two Applications

More extreme, frequent and longer heat waves negatively affect people all around the world, and especially inhabitants of urban areas which face even higher temperatures due to the urban heat island effect. A precondition to develop adaptation strategies to counteract adverse effects of heat in cities is to gain knowledge about the urban temperature distribution. One approach that has been applied in various cities is the implementation of dense urban temperature measurement networks. Since financial resources are usually limited, such networks consist of cost-effective measurement devices whose (daytime) data quality is prone to errors due to radiative influences. This was also the case in Zürich, Switzerland, where an urban temperature network with 272 measurement stations was operated from 2019 to 2021. In this study, we present a radiation correction method to enhance the data quality for practical use. Applying the proposed correction method led to a reduction in mean RMSE from 1.47 K to 0.57 K and in the overall mean bias from +0.88 K to +0.04 K. Following from that, we use the corrected database for two application cases: i) As a spatially and temporally high-resolution validation dataset for the physics-based large eddy simulation model PALM and ii) as input data for a geostatistical land use regression model. The analysis shows that the daytime radiation correction is crucial to detect the negative bias of the PALM model, which is most pronounced in the highly built-up area of Zürich, and to enhance the quality of the daytime land use regression. The developed radiation correction presented in this study can also be applied for other urban temperature networks that are facing similar challenges.