Crowdsourcing air temperature data for the evaluation of the urban microscale model PALM—A case study in central Europe

Abstract

In summertime and during heat events the urban heat island can negatively impact human health in urban areas. In the context of climate change, climate adaptation receives more attention in urban planning. Microscale urban climate modelling can identify risk areas and evaluate adaptation strategies. Concurrently, evaluating the model results with observational data is essential. So far, model evaluation is mostly limited to short-term field campaigns or a small number of stations. This study uses novel crowdsourcing data from Netatmo citizen weather stations (CWS) to evaluate the urban microscale model PALM for a hot day (Tmax ≥ 30°C) in Bochum in western Germany with anticyclonic atmospheric conditions. Urban-rural air temperature differences are represented by the model. A quality control procedure is applied to the crowdsourced data prior to evaluation. The comparison between the model and the crowdsourced air temperature data reveals a good model performance with a high coefficient of determination (R2) of 0.86 to 0.88 and a root mean squared error (RMSE) around 2 K. Model accuracy shows a temporal pattern and night-time air temperatures during the night are underestimated by the model, likely due to unresolved cloud cover. The crowdsourced air temperature data proved valuable for model evaluation due to the high number of stations within urban areas. Nevertheless, weaknesses related to data quality such as radiation errors must be considered during model evaluation and only the information derived from multiple stations is suitable for model evaluation. The procedure presented here can easily be transferred to planning processes as the model and the crowdsourced air temperature data are freely available. This can contribute to making informed decisions for climate adaptation in urban areas.