Study On Thermal Enviroment Of The Medium-sized City Using Multi-source Remote Sensing Data

Urbanization is the background and root of regional climate and ecologicalenvironment change, which results in numerous ecology and environment problemssuch as atmospheric pollution, ecology unbalance, abnormal of urban thermalenvironment and so on. And urban thermal environment is an increasingly prominentissue, which will become a serious impediment for urban sustainable developmentand residents’ living quality improving in the future. Aiming at these problems, thispaper takes the western medium-sized city—Mianyang city in Sichuan province as thetypical case, it reports the space–time evolution laws of urban thermal environmentunder the different time scales of inter-annual, season and diurnal based on the GIStechnology and multi-source remote sensing data, and also analyses the thermalenvironment effect of different urban landscape quantitatively. Main researches of thispaper are the following.(1) The retrieval models of land surface temperature are commented and chosen,which provide data support for analysis thermal environment.Land surface temperature is an effective indicator for charactering the thermalenvironment situation, its retrieval precision will decide the dependability of analysisresults directly. After comparative analysis the mainstream models of land surfacetemperature retrieval for the remote sensing data of AVHRR and TM/ETM+, theauthor selects the split window algorithm and algorithm based on imagery to retrieveland surface temperature for AVHRR and TM/ETM+data.(2) The evolution characteristics of urban thermal environment duringinter-annual, season and diurnal are reported by using Landsat TM/ETM+andNOAA/AVHRR remote sensing data.â‘ This paper takes TM/ETM+remote sensing images as data source, and useslandscape ecology theories and mean-standard deviation method to classify thethermodynamic landscape into five patterns. Then the author selects the landscapepattern indexes to count and analyse the change of landscape patches during inter- annual (1988~2011) and season. The thermodynamic landscape is destroyed seriouslyduring1988and2001, but this situation gets eased during2001and2011, which arereflected by four indexes including landscape fragment and so on. Furthermore, thedestruction extent of thermodynamic landscape is more serious in winter than spring,but better than autumn.â‘¡The changes of urban centroid and thermal centroid are all researched bybuilding the models of diversion distance and diversion direction on the centroidtheory. The results show that urban centroid and thermal centroid have stronglycorrelation with inter-annual change, during23years, the diversion direction of urbancentroid is northwestâ†'southwestâ†'southeast, the diversion direction of hightemperature patches centroid is northwestâ†'westâ†'southeast, the diversion directionof sub-high temperature patches centroid is northwestâ†'southwestâ†'southwest.Therefore urban expansion is the main driving force for changes of urban thermalenvironment. The various laws of thermal centroid among spring, autumn and winterare high temperature patches from northeast to southeast and sub-high temperaturepatches from west to southwest.â‘¢The Urban Heat Island Intensity(UHII) in May of four years(1988,2001,2007,2011) are3.65℃,1.77℃,1.07℃and0.55℃, which are computed by the methods ofurban-Rural mean temperature contrast and heat island area index. It is suggested thatthe Urban Heat Island Effect trend of Mianyang city becomes weaker and weaker inMay during23years. The UHII of Spring, autumn and winter are1.77℃,0.78℃and0.94℃, in this period the winter’s UHII is weaker than spring, but stronger thanautumn.â‘£After the profile analysis along W-E and N-S directions, the author finds thatthe relationships between land surface temperature and urban underlying surfaceproperties, population density, urban function partition are closely. The temperaturesof water and green are lower.⑤This paper takes AVHRR remote sensing images as data source, and useslandscape ecology theories and density segmentation to classify the thermodynamiclandscape into eight patterns, which also analyses the changes in diurnal of urbanthermal distribution and UHII. The results show that many thermal centers are locatedin Yongxing town, Gaoxin area, Fucheng area or economic-technologicaldevelopment area in the morning, noon and afternoon. Only one thermal center islocated in Fucheng area in the evening, night and early morning, the area of thermalcenters are different. The water’s thermal inertia is too large, so the temperature ofFujiang river and Anchang river increase slowly in the daytime, two rivers cut apartthe thermal centers. The temperature of Fujiang river and Anchang river reduceslowly at night, they impel thermal centers to be a whole. UHII is stronger at night than daytime, and the UHII from high to low are early morning, night, evening,afternoon, morning in one day,(3) The thermal environment effect of stream corridors, urban greenbelts andurban parks three typical urban landscape are researched.The thermal environment effect of stream corridors landscape, urban greenbeltslandscape and urban parks landscape are analysed. In brief, all of them have coolingfunction apparently, but the cooling effect is different. The best cooling effect of thethree is stream. The negative correlation between land surface temperature and NDVIis significantly except the stream by the profile analysis. Statistics of the averagetemperature of landscape pattern patch is that the temperature of woodland is thelowest and the temperature of the road is the highest of all. The relationships betweenthe cooling effect of urban greenbelts and area, perimeter, shape index are in directproportion. The author finds that the best cooling effect of greenbelts landscape area isabout2hm2by regression analysis the landscape area and average temperaturedifference. The shape is more complex, the cooling effect is more obvious. Thecooling effect of urban parks is not only related to the landscape area but also hasrelationship with landscape pattern. The urban parks with reasonable gradation amongarbor, bush, lawn and so on have better cooling effect.(4) Driving mechanism and countermeasure of urban thermal environmentchange are discussedUrban underlying surface properties, human heat emission and pollution are thethree main reasons which result in urban thermal environment change. In order toimprove residents’ living quality and thermal environment, we have to control theurbanization speed, protect natural vegetation, advocate low-carbon life, reduce theemissions of human heat and polluted gas.