| Land surface temperature(LST)is an important environmental parameter to characterize climate change,and plays an important role in the characterization of surface energy conversion.At present,there is no thermal infrared band in the high-resolution remote sensing image,but the spatial resolution of the thermal infrared image is not enough.In view of the contradiction between time resolution and space resolution,a method of surface temperature inversion using downscaling is proposed.In order to solve the above problems,based on the domestic high resolution two,fast bird multispectral image and Landsat-7 thermal infrared image,this paper adopts the mixed pixel decomposition method to carry out the method and Application Research of spatial scale reduction of surface temperature.The main conclusions are as follows:(1)Based on Landsat-7 image,the surface temperature is inversed by radiative transfer equation method,and the specific emissivity of the surface is calculated with a resolution of 30 m,and the surface temperature of 30 m is obtained by downscaling with EM algorithm.It is found that great changes have taken place in urban thermal environment from 2005 to 2015.The maximum value of surface temperature in 2005 is 55.65 ?,and the average value is 42.51 ?;the maximum value of surface temperature in 2015 is 65.05 ?,and the average value is 44.87 ?.(2)Based on the VHR-SUTM method and Distrad algorithm,combined with the GF-2 multispectral image of 4m and the quick bird multispectral image of 2.4 m,the spatial downscaling of 30 m surface temperature is carried out.When using distrade algorithm to fit the relationship between NDVI and LST,R2 is 0.391,which is far less than the ideal correlation coefficient value,so it is not suitable for this research area.Therefore,VHR-SUTM method is used to obtain high-resolution surface temperature,and quantiative and qualitative methods are used to test the accuracy.In 2005 and 2015,RMSE values of these two phases are 1.981 ? and 1.687 ?,respectively.In terms of visual effect,there is no significant difference between the reduced image and the original image.In general,the reduced effect is good,which meets the research requirements.(3)The temporal and spatial variation of surface temperature and the influence of different underlying surfaces on the surface temperature are studied.Through the mean standard deviation method to classify the thermal environment of the surface temperature,it is found that in 2005,the heat island area of Jinhua City was mainly located in the central region,and the highest temperature was only 57.11 ?,while in 2015,the heat island area of Jinhua city expanded to the surrounding areas of southwest and northeast,and the highest temperature reached 66.98 ?.It can be seen that in the past decade,not only the surface temperature of Jinhua City has increased,but also the heat island center has shifted.The scope of heat island is further expanded.This is mainly due to the change of building materials and the increase of urban greening in the central area of Jinhua City,which leads to the improvement of urban thermal environment in the central urban area,while a large number of forging factories are added in the southwest and northeast parts of Jinhua City,which leads to the deterioration of thermal environment in these areas.(4)Using the software of e Cognition,the paper classifies the data of GF-2 and Quick bird into object-oriented cart decision tree,and classifies the types of land objects into water body,building,cultivated land,forest land,grassland,road,unused land and other land objects.After classification,the surface coverage type image is superposed with the acquired downscaling temperature,and the occupied area and average temperature of each kind of surface feature are calculated.It can be found that from 2005 to 2015,not only the surface feature type has changed greatly,but also the surface temperature is significantly different.The cultivated land area decreased from 31.35% to 18.20%,and the surface temperature increased from 39.89 ? to 42.66 ?.the grassland area decreased from 21.70% to 16.52,and the surface temperature increased nearly 3 ?.the construction land area increased from 18.48% to 29.14%,and the surface temperature increased from 43.86 ? to 52.45 ?.The area of other soil species has little change,and the surface temperature has increased in addition to water.(5)In order to study the relationship between the density of different underlying surfaces and the average surface temperature,a multi ring buffer zone is established in the geometric center of the city.It is found that the fluctuation of the average temperature of buildings and roads is almost the same as that of the surface.Through the construction of scatter map,it is found that there is a positive correlation between buildings and roads on the surface temperature.The higher the density is,the greater the surface temperature will be,which will aggravate the urban heat island effect.Therefore,reasonable planning of urban buildings and effective control of bare land area are one of the methods to control the urban heat island. |
PDF Full Text Request