Abstract: Urban heat island effect is a very common and serious problem in modern city. It affects not only the quality of urban environment, but also the daily life of citizen. There are generally two methods to study it. One is to compare the meteorological data from the weather stations located in a zone between city and rural area, and another is to analyze the remote sensing images. Xi[JP8]’a[JP]n is a very important city in west China. In recent 30 years, many environment problems were resulted in along with the rapid urbanization, such as the reduction of farmland, soil and air pollution and urban heat island effect. In this paper, a case study was carried out in Xi[JP8]’a[JP]n City so as to retrieve the values of relative brightness temperature from Landsat TM thermalinfrared data (August 9, 2006 and August 4, 2010). The stability of relative brightness temperature during the period from 2006 to 2010 was analyzed according to the transition probability matrix of relative brightness temperature. The results showed that the brightness temperature intensity in Xi[JP8]’a[JP]n City was increased during the period from 2006 to 2010. The mean surface temperature was 26.28 ℃ in 2006, but was increased to 28.14 ℃ in 2010, and the increase occurred mainly in the western, southwestern and southeastern parts of Xi[JP8]’a[JP]n City. Based on the results of the relative brightness temperature, it was found that the extremely heat island effect did not occur in 2006 and 2010, the slight heat island effect was dominant, and the proportion of area of the patches with slight heat island effect was 61.11% and 58.64% in these two years respectively. The proportion of area of green islands was secondary, and it was 34.08% and 35.98% in 2006 and 2010 respectively. From the view of landscape ecology, stability of the patches with slight relative brightness temperature and of the green island patches was the highest. The stability of patches with moderate and severe relative brightness temperature was the lowest. It was considered that climate change and urban construction were the main reasons causing a dynamic change of surface heat regime. Urban green land could reduce the urban heat island effect.

Key words: urban heat island, Landsat TM image, remote sensing inversion, brightness temperature, thermal landscape, Xi’an