Research On The Relationship Between The Thermal Characteristic Of Underlaying Surface And Urban Heat Island

By using the remote sensing and GIS spatial analysis technology, this paper focuses on the monitoring and analyzing of urban heat island (UHI) effect in the old city of Beijing from combined ASTER and TM data. The land surface temperature of ASTER data was determined using an iterative self-consistent approach which is based on the split-window method. The land surface temperature of TM data was determined using common single channel algorithm. The heat island intensity was calculated by relative surface temperature. The urban heat island intensity was assessed by the relative land surface temperature which is defined asΔT/Ta, whereΔT represents the difference between the land surface temperature of urban area and the average surface temperature of country side, Ta is the average surface temperature of country side. Based on the heat island intensity, we investigated the spatial and temporal evolvement of UHI,the development of UHI along with the seasonal development and the UHI effect at night in the old city of Beijing. The three urban corridor structures-road system,greenbelt system and water system were used to do our research. The visible and near infrared channels of ASTER data were used to classification to obtain the underlaying surface types. The apparent thermal inertia computed from a real thermal inertia model, was used to analyze the correlation between the thermal inertia and the urban heat island intensity. Finally, the regression models were established. The conclusions of this paper were represented as follows:1. The UHI effect is obvious in the old city of Beijing and the heat island shows an enhanced trend with the development of urbanization. UHI effect is also changed with the seasonal development. At night, whether in winter or autumn, the temperature of the urban area is higher than country side which indicates a significant UHI effect.2. Generally speaking, these corridor structures have different effects on the urban heat island as follows. The cooling effect of water is the most obvious, greenbelt takes second place, the last system is road system whose direction is crucial to the alleviation of the urban heat island. Moreover the impacting range of the three corridor structures is limited, and usually not exceeding 300 meters.3. The visible and near infrared channels of ASTER data were used to classification using the decision tree,supervised and unsupervised methods. The underlaying surface types obtained from the classification were water,forest,grassland,road and building. It is discovered that the thermal inertias of water,vegetation are higher than the thermal inertias of road,building after the comparison analysis on the thermal inertia of underlaying surfaces which come from the classification image. The research calculated the correlation coefficients between the thermal inertia and the urban heat island intensity of five underlaying surfaces. The correlation coefficients are negative which mean the urban heat island intensity and the thermal inertia have a relationship of strong inverse correlation. The correlation degree is different at the different times and underlaying surfaces. In the 27th Jan ,2004, the highest negative correlation was found in fores(t-0.3970), followed by road(-0.3549)and water(-0.3140), and with a decrease for building(-0.2941). the lowest correlation was observed in grassland(-0.2883); However, the strongest, negative correlation, in the 31st Aug , existed in water(-0.5683). The correlation coefficient values dropped slightly for forest(-0.4882)and grassland(-0.4697), with a sharp decrease for building(-0.1811). The least correlation was found in road(-0.1403). The linear regression models between the thermal inertia and heat island intensity of various underlaying surfaces were established by a linear regression analysis. It was considered that the linear regression relation of the models were significant from the significance tests.