A landscape connectivity model to quantify contributions of heat sources and sinks in urban regions

The effect of landscape configuration on urban temperatures is always an important issue in landscape planning and mitigation of urban heat islands. However, landscape indices used in previous studies did not focus on thermal processes. We proposed a landscape source-sink distance (LSSD) index used to quantify the landscape connectivity and investigate its contribution to variations in land surface temperature (LST) in Beijing. Monthly LST was derived from MODIS remote sensing products in 2002 and 2012. Landscape composition and connectivity was calculated based on QuickBird and IKONOS images. The LSSD of each LST grid was calculated according to the accumulative shortest distance between green-impervious, water-impervious, and green-water types. The contributions of landscape composition and connectivity to variations in LST were assessed using a geographically weighted regression model. Heat sources and sinks were designated as having positive and negative effects on the LST, respectively. Results showed that (1) green spaces served as heat sinks both day and night. Water areas served as daytime heat sinks and nighttime heat sources; (2) the influence of green and water types on daytime LST varied in different months while their influence on nighttime LST was stable seasonally; and (3) a large distance between green and impervious land increased variations in day-night LST while a large distance for water-impervious connectivity might mitigate diurnal variations in LST. This study shows that landscape planners need to rationally use landscape connectivity among different landscape types and should focus on specific time and season for the effective mitigation of urban heating.