Exploring Urban Land Cover Changes and the Effect on Nocturnal Air Temperature Dynamics in Helsingborg - A temperature modeling made with TAPM

Abstract

In light of increasing global temperatures, urban areas face growing challenges related to heat stress. Alterations in land cover (LC) within cities contribute to changes in the intra-urban climate, predominantly driven by the replacement of vegetated surfaces with impermeable materials. These changes affect the thermal properties of urban environments, exacerbating the Urban Heat Island effect. Understanding the impact of land cover changes (LCC) on urban climate necessitates the use of remote sensing, pre-trained deep learning models, and temperature modeling techniques. This study focuses on assessing land cover changes in the Helsingborg urban area from 2004 to 2020 through image classification utilizing a pre-trained deep learning model. Furthermore, it investigates the influence of LCC on nocturnal air temperature using the three-dimensional prognostic air pollution and meteorological model, The Air Pollution Model (TAPM). Specifically, the analysis centers around the 2018 heatwave in Sweden, aiming to evaluate TAPM's ability to differentiate between various land cover types and identify temperature patterns within the Intra-Urban Heat Islands. The deep learning model achieved an overall accuracy exceeding 90%, revealing a decline in grass surfaces and an increase in areas covered by buildings and trees. TAPM's temperature modeling, based on the land cover classifications, demonstrated distinct temperature variations at a 100 x 100-meter local scale. Additionally, it indicated a higher proportion of areas with elevated nighttime temperatures (>18°C), posing potential health risks during heatwave events akin to the summer of 2018