The datasets are raster map layers produced using a linear regression model. Both map layers have been generated for the Turku region. In addition to Turku, the dataset covers some parts of the neighboring municipalities. The regression model for the modelled air temperature dataset (Turku_keskilampotila…) has been calibrated for the Turku area with the regression model, in which the response variable is the air temperature measured by the TURCLIM observation network (https://sites.utu.fi/turclim/). In the heat sum dataset (Turku_lamposumma…), the response variable has been the amount of degree hours exceeding 25 degrees Celsius. The explanatory variables of both models include the variables related to water areas, relative elevation, and urban land cover. Part of those datasets are available as separate datasets in Avoindata.fi site. Each situation's modeling has been employed with the variable-specific most appropriate radius for the explanatory variables, determined through correlation analysis and subjective considerations based on the theoretical background of urban climatology. The spatial resolution of temperature data is 100 m, and that of the heat sum 250 m. The temperature data cover land area, whereas the heat sum data cover land and water areas.
The data is produced as part of the HERCULES consortium/project funded by the Academy of Finland’s Climate Change and Health (CLIHE) research program implemented during the years 2020-2023.