Page 42 - 《华中农业大学学报》2026年第1期

36 华 中 农 业 大 学 学 报 第 45 卷

Effects of patterns of blue-green spatial landscape on
land surface temperature under different environmental

contexts in Wuhan City

XIE Qijiao,XIAO Shengyong
School of Tourism， Hubei University， Wuhan 430062，China

Abstract The quantitative relationship between all commonly used landscape metrics and land sur‐
face temperature （LST） across three spatial contexts including the main urban area， the urban develop‐
ment area， and the entire municipality of Wuhan City was analyzed by deriving the LST values and classify‐
ing land cover categories of Wuhan city based on Landsat 8-9 remote sensing data acquired on September
18 and 19， 2022 to study the effects of the patterns of blue-green spatial landscape on the LST to better
utilize the thermal environment regulation function of blue-green infrastructure. The principal component re‐
gression analysis was used to identify the dominant factors affecting LST under different spatial contexts
and reveal their underlying mechanisms. The results showed that water bodies and green spaces had a signif‐
icant “cooling island effect”， with the cooling intensity of water bodies （8.96-9.34 ℃） significantly greater
than that of green spaces （4.44-5.47 ℃）. Overall， the independent explanatory power of the landscape met‐
rics for LST changes followed in the order of water bodies > green spaces， landscape composition > spa‐
tial configuration， patch-level > landscape-level > class-level， and the main urban area > the urban de‐
velopment area > the administrative area. The dominant factors affecting LST varied across spatial con‐
texts. The four key factors in the main urban area were the percentage of water body area （PLAND_W），
water body patch density （PD_W）， effective mesh size of green spaces （MESH_G）， and edge density of
green spaces （ED_G）， collectively explaining 82.4% of the LST variation. The dominant factors in the
metropolitan development area were contrast-weighted edge density of water bodies （CWED_W）， percent‐
age of water body area （PLAND_W）， mean proximity index of green spaces （SIMI_MN_G）， and per‐
centage of green space area （PLAND_G）， collectively explaining 59.2% of the LST variation. The five
dominant landscape metrics related to blue-green spaces in the entire municipality， only explained 35% of
the LST variation. Water bodies and construction land had a strong explanatory power for changes in the
thermal environment， and the cooling effect of green spaces was significantly weakened or suppressed when
considering the combined effects of other landscape elements outside of blue-green spaces. It is indicated
that the regulation function of blue-green infrastructure in the thermal environment has a distinct context ef‐
fect. Measures for optimizing the spatial allocation and structural configuration of blue-green landscapes ac‐
cording to different environmental matrices including preserving large water bodies in central urban areas，
ensuring adequate water surface coverage， enhancing connectivity between smaller water bodies， enriching
the morphological complexity of blue-green spaces in urban development zones and metropolitan regions，
and strengthening their interactive frequency with surrounding environments can effectively enhance the
cooling performance of blue-green infrastructure.
Keywords blue-green infrastructure； heat island effect； cooling island effect； pattern of landscape；
context effect
（责任编辑：张志钰）