Observed & modelled nocturnal air temperature variations within an urban setting in Gothenburg. - and analysis of air temperature modelling with The Air Pollution Model (TAPM)

Abstract

Climate change is one of the most pressing concerns of the present day with numerous consequences arising from it, one of the most notably is rising air temperatures. An increasing urbanization and population growth in urban areas has become a key factor in this development, due to a greater expansion of impermeable materials, replacing the natural landscape, which effects the thermal properties of urban spaces. This also means that the effects of climate change have particularly strong effects on cities and its inhabitants. Rising air temperature throughout cities also leads to more people becoming vulnerable to its effects with an increased number of heat related health problems leading to hospital admissions and deaths, asserting an increased pressure on the population. Accompanying these developments is an exacerbation of the Urban Heat Island effect (UHI), meaning that urban areas become warmer than their urban counterparts at a higher intensity and rate. Understanding how different land uses affect air temperature in urban environments and how this will look like in the future, necessitates in-situ observations from different urban spaces and the use of meteorological models using higher resolutions. This study focuses on comparing the effect different land uses have on nocturnal air temperature in an urban setting during warm weather conditions in Gothenburg, by using in-situ observations taken during the summer months of 2023. Furthermore, it investigates how well the three-dimensional meteorological model, The Air Pollution Model (TAPM), can simulate these temperatures at sites located in built up areas classified as urban high. Two study areas and heatwaves were chosen, one at Gårda 17/6-26/6-2023 and one at Heden during 8/7-13/7-1999. Areas classified as urban high in Gårda displayed the highest minimum nocturnal temperatures compared to sites classified as urban low exhibiting the lowest with a difference of >=1°C most nights. A similar trend was found during the period in Heden where urban low indicated the lowest nocturnal temperatures and urban high the highest with a difference reaching >1.5°C. Internal variations at urban high sites were also exhibited during both periods with an average max temperature difference of 0.9°C at Gårda and >1°C at Heden. The meteorological model underestimated nocturnal temperatures in Gårda, achieving an R2 value of 0.15, indicating low correlation. During the Heden period the model overestimated the air temperature but achieved a R2 value of 0,60 indicating a substantially higher correlation. In Gårda, the model showed higher cooling rates when compared to the observations while in Heden the cooling rate where much more similar between observations and simulated. Comparing both study areas cooling rate indicated higher cooling rate at Heden than at Gårda. Indications of an intra-urban thermal breeze (IUTB) in Heden but not in Gårda, different Sky-View Factors and the size & structure of the open areas could provide explanation to the results produced by the model.