Analysis of urban climate based on earth observation and crowdsourcing data

Kurzfassung

Air temperatures obtained by point measurements at 2 m above the ground are essential for assessing citizen health and wellbeing issues related to heat stress, which are of major societal relevance, especially in cities. However, the paucity of official meteorological sites in cities prevents spatial information of air temperatures and the targeted implementation of heat mitigating countermeasures. To increase the spatial resolution of air temperature measurements, this study aimed at evaluating the suitability of using crowdsourced air temperature data collected from citizen weather stations (CWS) for climate analyses. Moreover, relations between air temperatures and land surface temperatures, land cover, and urban morphology parameters (building density, mean building height, and building volume) were investigated. These variables are available in full spatial coverage as they are derived from remote sensing products. The analyses were conducted for a study area in and around the city of Dresden, Saxony, Germany. Performing comparisons of quality-controlled air temperature records originating from the CWS and the official measurement sites, crowdsourced data were found to be suitable for climate analyses, although uncertainties about the representativeness and accuracies of CWS remained. Daytime land surface temperatures and coincident hourly air temperatures showed positive and moderate to strong correlations depending on the season. In the summer season, variations of land surface temperatures accounted the least for variations in air temperatures. Analyzing the characteristics of summer temperatures according to land cover types, it was found that on-site (i.e., at the location of temperature measurements) land cover was a good indicator of air temperatures. Artificial surfaces revealed the highest median temperatures, high vegetation sites the lowest. Investigations of fractions of land cover types within a given surrounding allowed a more detailed differentiation of temperatures. The examined urban morphology parameters generally showed positive relations with air temperatures but revealed unsteady trends. The results of this study contribute to a better understanding of the spatial distribution of air temperatures in relation to variables that characterize the earth’s surface and are available in higher spatial resolution. Finally, due to the influence of the investigated parameters on air temperatures, these parameters should be considered in models estimating air temperatures.