The Urban Heat Island Analysis of Changsha-zhuzhou-xiangtan Urban Agglomeration Aased on Modis Data

Using a spatial resolution of MODIS land 1000m standard products ,we can get the Land Surface Temperature.Researching for the Land Surface Temperature including spatial and temporal distribution characteristics influence factors.The results show that Spring,Summer and Autumn temperatures mainly concentrated in the central region,Winter temperature mainly concentrated in the South region.From 2001 to 2015,the maximum temperature difference is summer daytime and the difference is 17.58°C,the minimum temperature difference is autumn daytime and the difference is 11.3°C.According to the thermal field intensity distribution,compared 2005 with 2015,Urban Heat Island intensity gradually increased in 2015,the high temperature area increased and distributed more concentrated,and diffusion weakened from the city to the surrounding,the urban heat field is higher than the thermal field.That index by calculating the thermal landscape,account for a dominant position in the middle of heat distribution,and all types index in 2015 are higher than in 2005.


Introduction
With the intensification of urbanization and industrialization, the area and scope of urban construction land is expanding, which makes the city have a phenomenon that the urban temperature is higher than the suburban temperature，called the "heat island effect", it is affected by the urban landscape type and the impact of urban spatial patterns [1][2][3][4] .

Data collection
The MODIS data used was downloaded from the LAADS Web [5] .The data of the four seasons of 2001, 2005, 2010 and 2015 were selected to analyze the seasonal and interannual variations of urban surface temperature in the Changsha, Zhuzhou and Xiangtan areas in the past 15 years, and the distribution of urban island heat island characteristics was obtained.

Data preprocessing
After obtaining the remote sensing image, it is first subjected to radiation calibration, atmospheric correction and other pretreatment;Then, the monthly maximum, seasonal average, and annual average data are synthesized for the surface temperature data, and the composite results are shown in Fig. 1.

Analysis of interannual variation of surface temperature
Table 2 shows the statistically obtained four-year interannual average temperature data and four-day average surface temperature information.In order to obtain the variation of the surface temperature between years, the difference calculation is compared, and the result is shown in Fig. 3.

Thermal field strength analysis
Heat island intensity is one of the important indicators for heat island effect assessment [7] .In order to improve the comparability of remote sensing images,the temperature image maps of 2001, 2005, 2010, and 2015 were normalized, and the normalized formula was [8]

Thermal landscape pattern analysis
Common characterization methods for landscape pattern characteristics include landscape pattern index and spatial statistical methods [9][10] .Select the following landscape index to quantify the characteristics.

Conclusion and discussion
Through the research in this paper, we can draw the following conclusions: According to the four-year average temperature map of the study area, the thermal field intensity in spring, summer and autumn is weakened from the center of the city to the periphery, and the trend from the north to the south is weakened and then weakened.The winter high temperature area is mainly concentrated in the central and southern regions.The thermal field strength is the strongest.

Fig. 2
Fig. 2 calculates the anomalies for each season based on the daytime average of each season, and the difference results can reflect the difference in surface temperature in the same season in different years.
Density segmentation of normalized results.The classification results are shown in Fig. 4.

3.1 Analysis of temporal and spatial variation of surface temperature
in 2001, 2005, 2010, and 2015, the seasonal average of the corresponding four years, and the seasonal average and range of values for the four seasons are obtained, as shown in Table 1.E3S Web of Conferences 53, 03045 (2018) https://doi.org/10.1051/e3sconf/20185303045ICAEER 2018

Table 1 .
Seasonal mean and range of values.

Table 4 .
Dynamic change of plaque type level index

Table 5 .
Dynamic change of landscape level index.