南京市热环境与地表覆被的时空尺度效应及驱动机制研究

Spatio-temporal Scale Effect and Driving Mechanism of Thermal Environment and Land Surface Cover in Nanjing

近年来城市热环境与地表覆被之间的时空尺度效应研究备受人们关注。利用2017年四期Landsat 8遥感影像提取南京市四季地表温度与主要地表特征参数,从30—16000 m的14个网格尺度上研究了地表温度与地表特征参数的相关关系,并基于此,探讨了南京市热环境的时空尺度效应及驱动机制。结果表明,南京市地表温度与地表特征参数的相关性具有不同的时空尺度效应。归一化植被指数(Normalized Difference Vegetation Index,NDVI)在春、夏、秋三季与地表温度具有显著相关性的最大研究尺度分别为300、12000、8000 m;归一化建筑指数(Normalized Difference Built-up Index,NDBI)在春、夏、秋三季与地表温度具有显著相关性的最大研究尺度则分别为15000、16000、6000 m;改进的归一化水体指数(Modified Normalized Difference Water Index,MNDWI)与地表温度具有显著相关性的最大研究尺度在春、夏、秋三季均为12000 m。南京市冬季NDVI、NDBI及MNDWI这3种地表特征参数在不同网格尺度上与地表温度的相关关系均较弱。季节因素较尺度因素对地表参数与地表温度关系的影响更为显著。春、夏、秋三季NDBI与地表温度呈显著正相关(P<0.01),且增温效应表现为夏季>春季>秋季。夏季无论何种网格尺度,测定系数R2均表现为NDBI>NDVI>MNDWI,且NDBI线性回归的斜率均远大于NDVI和MNDWI,表明建筑等不透水面是南京市夏季城市热环境恶化的主要驱动因素。研究结果对南京市城市规划的合理布局,维持城市的可持续发展具有重要的参考价值。

The exploration of the spatial-temporal scale effect between the urban thermal environment and the land surface cover attracted much attention in recent years.The four seasons land surface temperature and main land surface characteristic parameters in Nanjing were extracted from four Landsat8 remote sensing images in 2017.The correlation between the land surface temperature and the land surface characteristic parameters was studied from 14 grid scales ranging from 30 m to 16000 m.The spatial-temporal scale effect and driving mechanism of the thermal environment in Nanjing city were discussed.The results showed that the correlation between land surface temperature and surface characteristic parameters in Nanjing had different spatial-temporal scale effects.The largest research scales for the significant correlation between the normalized difference vegetation index(NDVI),and the surface temperature in spring,summer,and autumn were 300,12000,and 8000 m,respectively.The largest research scales where the normalized difference built-up index(NDBI)had the significant correlation with the surface temperature in spring,summer,and autumn were 15000,16000 and 6000 m,respectively.The largest research scale for the modified normalized difference water index(MNDWI)significantly related to the surface temperature was 12000 m in spring,summer and autumn.The correlations between the three kinds of surface characteristic parameters and land surface temperature in Nanjing in winter were weak on different grid scales.Seasonal factors had more significant influence than scale factors on the relationship between surface parameters and land surface temperature.There was a significant positive correlation between NDBI and surface temperature in spring,summer and autumn(P<0.01),and the warming effect was summer>spring>autumn.At all grid scales in summer,the determination coefficient R2 were shown as:NDBI>NDVI>MNDWI,and the slopes of the linear regression of NDBI were much greater than that of NDVI and MNDWI,which indicated that impervious surfaces such as buildings were the main driving factor for the deterioration of urban thermal environment in Nanjing in summer.This work might provide important reference value for the rational layout of urban planning and maintaining the sustainable development of Nanjing.