The cooling effects of blue–green spaces on the urban heat island effect are complex and different. The purpose of this study is to simulate how the cooling effect of blue–green space changes with its size and shape. The cooling effects of 53 green patches and 28 water bodies in Changsha were extracted based on Landsat images. A surface fitting model was used to quantitatively reveal the relationship between the cooling effect of blue–green space and its size and shape. The results show that the cooling effects of blue–green spaces were enhanced with the increasing size, and then would become stable after a certain range (threshold). Certain thresholds were identified between the blue and green space areas (2.98 ha and 3.15 ha, respectively) and the cooling distance, and between the blue and green space areas (4.84 ha and 4.92 ha, respectively) and the cooling magnitude. In addition, the green space with an area of 9.08 ha and landscape shape index (LSI) of 2.97 could achieve a better cooling distance (413.46 m); and the blue space with an area of 29.4 ha and LSI of 1.75 could achieve a better cooling magnitude (5.17 °C). These findings provide useful guidelines for urban planning and improving urban livability in other regions with terrain and climate conditions similar to Changsha.

Urbanization has significantly changed thermal environments and vegetation phenology. However, the effects of spatially different land surface temperatures (LST) on vegetation phenology, rather than differences between urban areas and rural areas, remain unclear. In this study, four cities with similar vegetation types located in temperate monsoon climate zones were selected to map vegetation phenological metrics and discuss their responses to spatially heterogeneous LST within urban areas. First, Sentinel 2-A and 2-B data were used to estimate phenological metrics by combining Savitzky–Golay filtering, and Landsat 8 TIRS data was used to obtain LST. Second, buffer zones (from the urban center to the urban edge at 1 km intervals) were used to extract the averaged phenological metrics and LST. The response of the phenological metrics to LST from the urban center to the urban edge was then analyzed. Results show that spatial differences in LST and vegetation phenology exist inside urban regions as well as between urban and peri-urban areas. In addition, the response of phenology to LST within urban areas is also obvious. SOS is negatively related to spring LST from the urban center to the urban edge, whereas EOS is positively related to autumn LST.