Study On The Method Of Land Surface Temperature Downscaling And Its Application In Urban Heat Island

The acceleration of urbanization leads to the replacement of natural features by artificial surface,which changes the process of material circulation and energy exchange on the surface,and the resulting heat island effect has many negative effects on the ecological environment and human health.It is of great significance for urban planning and social development to strengthen the monitoring of urban heat island effect and the study of its genetic mechanism.However,in the past,the problems of high temporal and spatial resolution of land surface temperature were restricted by each other.In view of the above situation,this paper introduces the method of land surface temperature downscaling to make up for the lack of data,and applies it to the study of heat island effect.Then,the long-term fine scale monitoring and characterization of Beijing heat island are carried out,and the temporal and spatial correlation analysis of the driving factors of heat island is carried out.The main research results are as follows:(1)A method of land surface temperature downscaling based on multi-factor geographical weighted regression(MFGWR)is proposed.Multi-temporal Landsat 8 data of four research areas is used to calculate the environment covariates and the simulated coarse resolution surface temperature is downscaled.The algorithm is compared with other methods to evaluate its applicability.According to the analysis of error spatial distribution and uncertainty indices,the algorithm in this paper has higher precision in high heterogeneity area and is more suitable for the study of urban area because of considering multi surface parameters and geographical weight.(2)Based on sentinel 2 and HJ-1A / B data,the MODIS land surface temperature is downscaled,and a long time-series of 100 m surface temperature in summer from 1984 to 2019 in Beijing is generated by integrating multiple Landsat satellite images.The intensity of heat island is calculated by urban-rural difference method,and the intensity of heat island is graded by Getis-Ord-Gi* method.In view of the heat island area and position of information data,the grading results of heat island intensity is analyzed quantitatively and qualitatively: the evolution of heat island in Beijing spreads from the central urban area to the surrounding area,from fast to slow,with obvious local aggregation effect;the main causes of heat island include vegetation cover change,impermeable layer change,economic development and population expansion and other socio-economic factors.(3)From the two dimensions of time and space,the temporal and spatial correlations of heat island area,vegetation coverage change rate,impervious layer area and various socio-economic indicators were explored by using sens slope trend analysis,grid density analysis and correlation analysis.The interannual analysis shows that the correlation of heat island area with vegetation coverage and impervious layer area increases first,then decreases,and then fluctuates.The driving mechanism is becoming more and more complex.The analysis of districts shows that: the reduction of heat island area in central urban area is related to vegetation planning;the heat island effect in urban expansion area is greatly affected by per capita GDP and population expansion;the heat island effect in most new development areas is greatly affected by industrial development.For a small number of districts and counties with no obvious fitting effect,the driving factors are unstable.