Modelling the Atmospheric Urban Heat Island and its Contributing Spatial Characteristics

The Urban Heat Island (UHI) is a phenomenon in which a city forms microclimates which heat up significantly quicker than its rural surroundings. Urban environment has a tremendous and not extensively explored impact on human health and on environment. Until now the UHI was mostly studied by the means of satellite remote sensing techniques. This approach has four basic drawbacks that the current paper tackles. Firstly, remote sensing does not allow examining the air temperature (Ta) for specific days; for instance days with greater scientific interest (e.g. heat-wave phase). Secondly, the spatial resolution cannot be altered depending on the research needs. Moreover, satellite images (e.g. Landsat 8) enables only per 16 days updates which is not enough to be able to capture the UHI distribution dynamically during a heat-flux. Finally, the estimation of Ta by using satellite images is based on the surface temperature (Ts). However, this paper judges that the relation between the Ts and Ta is not always strong and positive. To overcome these limitations, this research utilizes weather stations to enable Ta measurements with hourly updates which enable diurnal and seasonal observations. This paper investigates a methodology to analyze and visualize dynamically the Atmospheric UHI. The method followed is reusable since it can be applied for other cities or other time spans for which all the necessary data is available. To calculate the location variables (independent variables) which contribute to the UHI, data from different sources were acquired and utilized. Moreover, a dataset derived from 140 sensors distributed in The Hague area constitutes the most important dataset since accompanied with the KNMI temperatures dataset enables the timely analysis and visualization. All the data were integrated and organized into a grid with 100m *100m resolution. Data modelling is accomplished for all grid-cells in the area of interest by using the relation between the sensed temperatures with the location variables. The sensed temperatures were further used to validate the constructed models. During this research we observed that Ta and Ts have different behavior. Ta depends on the rural temperature, albedo, coastline proximity, footprints, sky view factor, Normalized Difference Vegetation Index (NDVI) and imperviousness. However, there is no relation with water surfaces in contrast with Ts. Moreover, the relations are not as strong as with Ts. These findings highlight the need for extra research on this topic by examining variables that may affect more the Ta, such as the wind direction. Therefore, the final output of the paper is a dynamic representation of UHI during the heat-flux of 2015. The visualization aims to help urban planners, designers and policy makers to understand the microclimates in order to develop strategies and designs towards more liveable cities.Architecture and The Built EnvironmentGeomatics for the Built Environmen