URSUS_LST: URban SUStainability intelligent system for predicting the impact of urban green infrastructure on land surface temperatures

Abstract

Mitigating Urban Heat Island (UHI) effects has become a challenge to improve urban sustainability. The simulation tool URSUS_LST has been developed to allow urban planners to estimate how the addition of different green infrastructure elements would affect temperature. To achieve this, a new methodology was defined based on data mining, geospatial image processing and the knowledge of experts in the domain that predicts the Land Surface Temperature (LST) of any location within a city. It consists of a first data mining phase in which the real LST and the different urban elements of the nearby environment are considered: buildings, vegetation and water bodies. In a second phase, different regression models are induced to predict LST. Additionally, considering the most accurate models, the relevant attributes and their relationships are identified. A real application of the tool in the city of Malaga (Spain) has been used as an example of its usefulness.