Author:
Westerholt René,Klopfer Florian
Abstract
Abstract. Both the degree of urbanisation and climate change are continuing to accelerate. As a result, heat islands in cities are an increasing problem. In this article, we analyse a dataset of land surface temperatures in Berlin. The heat information is not mapped on administrative or census units, but summarised into so-called urban structure type units. These are not only meaningful in terms of their inherent link to urban heat but also represent a geography that is relevant for planning purposes. Our analysis complements existing studies in Berlin by applying two novel visual spatial analysis methods: the Moran drop plot and the Moran seismogram. These allow structures to be visually explored that are not found in traditional Moran scatterplots. In this way, we can visualise that low-temperature areas tend to be more homogeneously clustered than their high-temperature counterparts. In turn, the latter can be identified in the plots as clearly smaller in geographic scale and thus as more heterogeneous in the neighbourhoods formed. Finally, we draw conclusions including possible policy-relevant considerations.