山地城市坡向对地表温度的影响——以重庆市主城区为例

为揭示坡向对山地城市地表温度的影响,以重庆市主城区为例,由Landsat TM影像反演地表温度,提取各坡向上的地表温度,进行方差分析。结果发现,东南和南坡向的地表温度较高,而西北和北坡向的地表温度较低,8坡向分类时,各坡向的平均地表温度的差异具有显著性;用地类型没有坡向偏好,对于同一种用地类型,各坡向的平均地表温度没有显著差异。不同的用地类型上的城市建设强度和生产、生活活动差异是引起地表温度变化的主要因素,而坡向对城市地表温度的影响相比强烈的城市人为活动产生的环境效应而言是微弱的。

Influence of aspect on land surface temperature in mountainous city: A case study in central area of Chongqing city

In mountain areas, land surface temperature (LST) is a key parameter in the surface energy budget and is an outcome of a complex interplay of topography, incoming radiation, atmospheric processes, soil moisture, land covers and vegetation types. One important factor that controls LST is topography. Slope and aspect are generally considered as important parameters in ecology as they directly influence the potential radiation and heat load in small scale. In built-up areas in mountain city, where land use/cover is completely dominated by human activities that exert a great influence on the local climate, biogeochemical circle and landscape diversity, LST is influenced not only by topology but also by human activities. In order to reveal the influence of aspect on LST, central area of Chongqing city, a typical mountain city, has been chosen as an example. Land surface temperatures (LSTs) derived from Landsat TM images were extracted by aspect using spatial analysis function in ArcGIS, and were analyzed using analysis of variance (ANOVA) in SPSS. Firstly, the scatter diagram of aspect and LST, that were extracted for each pixel, shows neither simple linear nor nonlinear relationships. Secondly, the aspect is divided into different number of classes, namely, four classes (every 90°), eight classes (every 45°), 16 classes (every 22.5°), 32 classes (every 11.25°), 64 classes (every 5.625°), 128 classes (every 2.8125°) and 360 classes (every 1°). The results show that LSTs in the southeast aspect and the south aspect are higher than other aspects, whilst LSTs in the northwest aspect and the north aspect are lower. Although aspect is an important factor influencing earth's surface energy received from the sun radiation, using finer resolution in aspect makes no significant difference from using coarser resolution. Among all class divisions, when the aspect is divided into 8 classes, the average LSTs in each aspect have the most significant differences. The average LST of each aspect is positively correlated to its area proportion in the study area. The average LSTs in each land cover/use are significantly different, however, the average LST in each aspect class is not significantly different in the same land cover/use type. This implies that the LST distribution in city is mainly caused by urban construction and human activities, roughly corresponding to different types of land cover/use. Comparing to strong environmental effects resulted from intensive and large-scale urban construction activities, the influence of aspect on LST is relatively weak. In mountain city, urban construction activities massively transform original natural topography into artificial terraces and the urban construction size is much greater than the size of hill reserved inside the city, hence it is difficult to identify aspect effect on LST. The most likely reason for this is that the differences of light condition, soil moisture, precipitation and wind velocity result from aspect are disguised or offset by intensive human activities.