high performance computing on graphics processing units: hgpu.org

Metropolitan centers can be affected by an urban heat island effect. Radiative heat build-up from pavement and buildings increases temperatures in the metropolitan area above the average temperatures normally found in the surrounding environment. One way to help reduce the heat island effect is to add parks, trees, or green roofs to these urban spaces. Understanding the impact of these green spaces, including the amount and placement needed to best reduce the heat island, is not well understood. The primary objective of this thesis is to develop a fast and efficient simulation framework for modeling the heat transport in urban centers. These efforts are part of a larger environmental modeling and optimization system that is gathering knowledge about how urban structure and green infrastructure impact energy use and pollution dispersion in urban spaces. Our model uses a real-time ray tracing engine to support fast, radiative heat and energy transfer within urban settings. The framework is able to calculate temperatures in under 1.8 seconds with an average setup time of 1.84 seconds. This work also provides visualizations of the amount of incoming shortwave solar radiation, the amount to which each small area of surface is visible to the sky, and the temperatures on surfaces. These visualizations are provided in order to give insight into the understanding of complex environment and urban interactions.