Evaluation of the urban climate model PALM-4U with conventional and non-conventional monitoring data for a case study Vienna, Austria

The newly developed Parallelized Large-Eddy Simulation Model for Urban Applications (PALM-4U) has been applied for the city of Vienna to simulate thermal conditions in urban environment and to evaluate its performance using different modelling configurations as well as variable spatial scales.

Model simulations were carried out for a representative clear-sky heat day in August 2022 for the entire city of Vienna with a spatial resolution of 20 m and for a selected residential area in the city with a spatial resolution varying between 1 and 10 m. The data related to geographical information and urban structures are based on high-resolution GIS data provided by the city of Vienna, including multi-purpose area map, street tree cadastre, digital elevation and surface model, which were combined with the national land cover data (Land Information System Austria - LISA) to account for the unresolved vegetation and Open Street Map to include building properties in surrounding areas of the city. Selected model configurations enable evaluation of model performance both on micro-scale taking into account variable building types, resolved vegetation and detailed surface properties as well as on city-scale including interaction with regional weather conditions and complex terrain.

The results for hourly air temperature were compared with the conventional weather stations of the national weather service and the city of Vienna. In addition, quality-controlled data from more than 1000 citizen weather stations from the company Netatmo available in Vienna were used to compare differences in temperature variability and spatial patterns. The results show high intra-urban variability during daytime, but distinct spatial pattern during the night.

In order to evaluate the temperature variations on micro-scale, a measurement campaign at the headquarter of the Austrian weather service using 25 low-cost weather stations equipped with LoRaWAN Dragino LHT65 sensors was conducted. The stations were positioned in the vicinity of different artificial and natural surfaces (e.g. grass, parking lot, street) and structures (e.g. concrete and green roof) and the retrieved data were used to evaluate temperature variations in the high-resolution model setup. Both, measurements and the model results, show high variability in air temperature during daytime and higher temperatures near building structures than vegetation. Due to extremely dry conditions, natural surfaces, such as free soil or low grass show high daytime temperatures as well.