Dynamic Of Urban Heat Island(UHI) And The Responsive Mechanism Of Local Climate Zones To UHI In Fuzhou

Fuzhou,as a typical representative city with hot summer and warm winter,is known as"the first stove city".The high temperature in summer has become the primary factor affecting the livability of Fuzhou.At present,there are many researches on Urban Heat Island,most of which are based on the pixel scale of remote sensing data.Based on the Landsat remote sensing data,this paper uses the single window algorithm to retrieve the land surface temperature of the main urban area of Fuzhou from 2000 to 2020,and takes the urban block landscape unit as the research scale to explore the urban heat island dynamics of Fuzhou in recent 20 years.Taking 2020 as a representative,the driving mechanism of urban heat island is explored through the geographic detector model,and the technical scheme of local climate zone division and the response mechanism of heat island in Fuzhou are explored combined with urban morphology and land cover types.The main conclusions are as follows:(1)Heat island dynamics in Fuzhou in recent 20 years:the heat island intensity ranges from 0.10?to 10.00?in recent 20 years,and the heat island intensity of significant heat island areas(level-3 and level-4)are 3.33?-6.19?and 5.25?-10.00?respectively.The area of significant heat island increased from 121.29 km~2to 177.71 km~2in recent 20 years,which increased by 56.42 km~2(accounting for 17.72%of the study area);In the past 20 years,the total change area of heat island is 169.51 km~2,accounting for 53.23%,of which 128.92 km~2is deteriorated and 40.60 km~2is slowed down;In terms of spatial distribution,the spatial patch and density of the hot spot area(level-4)continue to increase,and its aggregation degree is relatively stable,forming a relatively stable hot spot space.(2)Driving mechanism of Fuzhou heat island:through the geographic detector model,the paper analyzes the four factors that may affect the urban heat island effect,including terrain,surface coverage,artificial heat and building intensity,with a total of 12related indicators,including DEM,MNDWI(improved normalized difference water body index),NDVI(normalized vegetation index),POI(point of interest),impervious surface,average surface area,average height,slope The driving mechanism of aspect,plot ratio,road density and building density on UHII.In the whole study area,the indicators and driving forces of UHII(urban heat island intensity)are ranked as follows:NDVI,impervious surface ratio>MNDWI>plot ratio>average height>slope>average building surface area>DEM>aspect.These indicators have a synergistic effect on UHII,and RF(random forest,The results showed that R~2,RMSE and RPD of the model were0.71,0.97 and 1.87,respectively.The nine driving indicators could reasonably predict the regional UHII.(3)LCZ(local climate zones)division and its response mechanism to heat island:16LCZ types in the study area can be divided according to the indexes of building coverage,building height,building density,NDVI value,MNDWI and impervious surface ratio.The total area of dense forest area(LCZ-A)and water body(LCZ-G)is 77.61 km~2(24.37%)and 30.22 km~2(9.49%),open high-rise(LCZ-4)and dense middle-level(LCZ-2)accounted for 14.56%and 9.01%respectively.The hot spots are mainly distributed in the dense middle layer(LCZ-2)and the open middle layer(LCZ-5).The proportion of level-4 is 68.61%and 46.43%respectively.The proportion of level-3 is84.55%,67.72%,77.88%and 75.59%respectively in the UHII of dense high-rise(LCZ-1),open high-rise(LCZ-4),open low-rise(LCZ-6)and light low-rise(LCZ-7).Each LCZ type can be predicted linearly with specific driving factors,and the heat island intensity of the hot spot area dense middle layer(LCZ-2)and open middle layer(LCZ-5)can be predicted linearly with NDVI,MNDWI and plot ratio,and the R~2is greater than0.66.