Temporal And Spatial Changes Of Urban Thermal Environment Remote Sensing Research In Shenyang City

Since the urban heat island effect was first proposed in1833, of which theinfluence on the enviroment has been gradually focused on. The optimization ofnatural and living environment has become a national strategy goal with thedevelopment of the urban economic construction. The urban thermal environment isthe comprehensive reflection of urban space environment in the thermal field. Thetime, spatial and spectral characteristics of remote sensing are the data basis ofbuilding the surface models and quantitative research. They can also support thedynamic monitoring and effected factor research of urban thermal environment. Thispaper takes the Shenyang City as study object to analyse the temporal and spatialvariation of urban thermal environment and provide a reference for urban ecologicalconstruction. The research achieved the results below:1. Dynamic monitoring of urban thermal environment on the study area havebeen realized. LandsatTM/ETM+image data of the year1992,2001,2010were takenas data resource and the single-window algorithm was used to invert the land surfacetemperature. Supervised classification combined with visual interpretation was usedfor landscape classification on the study area. Landscape pattern was quantitativelyanalysed by landscape indices, of which the effect on urban thermal environment wasstudied. The CA-Markov model suitable for study area was built for simulationprediction on study area’s thermal environment.2. Heat island changes of study area was analysed. The land surface temperatureinversion result was divided on basis of mean standard deviation method with surfacethermal field level diagram drawed, on which the area statistics was carried. Theresult was that the area percentage of high temperature zone. Increased from4.5%to12.8%. The range sprawlly developed outward along the center. The west and southhad obvious expansion. Compared to the year2001, high temperature zone of2010was divided into small area plaque from a large flake area and sub temperature ofmid-temperature zone, and the heat island intensity reduced from4.98to4.56; In thepast20years, heat island range has apparently expanded and heat island intensityincreased first and then reduced. 3. The surface landscape has an obvious correlation with thermal field level.Refered to the landscape classification map and surface temperature map, the averagetemperature of landscape types was calculated, whose order from high to low was:construction land> urban green land> arable land> woodland> water. The heat levelchange map of1992to2001and2001to2010was made. The enhanced level meantthat vegetation area and water changed to construction land and the unaltered levelment the landscape types didn’t change. Changes of surface heat level landscape wereapparently affected by the landscape change. Plaque index, boundary index and thediversity index were mainly chose for illustrating landscape patterns changce. Theresult showed that Green land and construction land’s fragmentation enhanced withheat island intensity reducing. Water and woodland aggregation index increased,which formed "cold island" area. Area fractal dimension and edge density and otherindicators were increased and the patch edge effects were enhanced, which wasconducive to the heat flow between patches. Connectivity and heterogeneity increased,which was convenient for energy exchange between landscapes. Landscape pattern’schange impacted on the distribution of the thermal environment.4. CA-Markov model suitable for thermal environment prediction was built. Inaccordance with transition probability matrix of surface heat field in the year2001and2010, suitability image sets were determined and CA-Markov heat field modelwas built.The real data of2010was used for verifying model’s accuracy. This modelcould be used to study the thermal environment simulation. Based on the2010thermal field scape graph,2019thermal environment in Shenyang City was predicted.Through the statistical analysis, high temperature showed a slightly increasing trend.but ares of high temperature and sub-temperature largly increased. The urban heatisland effect was reduced but still relatively obvious. Suggestions were recommendedseparatly from aspects of human activities and landscape pattern for urban thermalenvironment optimization.