Abstract
In terms of struggling against urban warming, increasing the cold island potentials is the easiest, least costly, and most environmentally friendly method to be taken into consideration in climate-sensitive urban design/planning studies. However, the difficulty of creating new cold islands in densely built urban areas without any urban transformation project is an important handicap in the application of the method. At this point, road networks draw attention as the most important spaces that can be used, especially in the built environment. Transforming these spaces into green streets by afforestation creates cool lines in densely built-up areas, provides an increase in urban climate comfort in their near peripheries and a decrease in energy consumption indirectly. In this study, the effect of increasing street afforestation on the local climate in Beşiktaş, one of the densely populated districts of Istanbul, was investigated by artificial neural network (ANN) simulations. In the method based on NDVI (normalized difference vegetation index) values obtained from satellite images, new synthetic images were created by afforesting road networks and then new surface temperatures of the region were calculated from the simulations that used these synthetic images. By analyzing the thermal changes between the obtained simulated surface temperatures and the original surface temperatures, the effects of street afforestation on the region were examined. The results revealed that street afforestation is also effective in cooling the road networks and their peripheries. Keywords: Street afforestation, Cool island, Urban climate change, Artificial neural network (ANN), Remote Sensing, Thermal change detection
