Spatial modeling of Zaragoza urban temperature at seasonal scale, 2015-2022.

The aim of this communication is to use Geographic Information Technologies (GIT)  to improve the characterization of the spatial variability of temperature in urban spaces. For this purpose, an interpolation method has been developed that allows to know the spatial patterns of the maximum, mean and minimum urban temperature of Zaragoza at seasonal scale. Using as a climate database the observatory network of the Clima, Agua, Cambio global y Sistemas Naturales research group, composed of 21 thermohygrometric sensors distributed in different urban areas of the city and its surroundings. We will use daily scale temperature data from March 2015 to December 2022 to apply them in a cokriging interpolation model, complemented with spatial information of different variables. These complementary variables of satellite and territorial origin will be subjected to a previous correlation and regression analysis with temperature, including in the interpolation only those significant variables with higher correlation, discarding those correlated with other covariates. For the evaluation of the model, different combinations of model, variables and principal components were calculated using different error metrics to check which was the most appropriate option. Finally, the first three principal components of the variables Sky View Factor, DEM, Green Leaf Index, Normalized Difference Turbidity Index, Normalized Built-up Area Index and Land Surface Temperature were chosen. Finally, we obtained as a result the spatial distribution of the temperature in the city of Zaragoza at 100 meters, a seasonal average interpolation of the study period. Obtaining a complete picture of the spatial distribution of urban temperature that allows analysis at detailed scales between different urban spaces.