Global urbanization and population growth have led to numerous environmental and ecological challenges. Specifically, the concentration of developed areas and increasing populations consistently worsen urban thermal conditions, culminating in significantly higher temperatures than those in suburban areas. This phenomenon is known as the urban heat island (UHI) effect. In the current work, we study the cooling effects of urban green spaces (UGS) defined as land that is partly or completely covered with vegetation, such as trees, shrubs, grass, or other plant life. By leveraging remote sensing data to identify UGSs and analyzing historical meteorological data (temperature, humidity, wind), this research aims to compare the UHI intensity of areas surrounding UGSs with those of non-UGS regions in the Houston metropolitan area. Convolutional Neural Networks (CNNs) will be employed to model and predict the UHI effect around UGSs and across other city areas lacking green spaces. The study will also forecast future UHI scenarios, based on potential changes in the size and distribution of UGS in both UGS and non-UGS areas. The findings will provide valuable insights for urban planners to optimize green space distribution for improved thermal regulation and urban resilience.