Research And Application Of Hourly Simulation Of Urban Surface Temperature In Daytime

The simulation study of the diurnal variation of urban surface temperature is of great significance to the monitoring and analysis of summer high-temperature heat waves,the spatial distribution and dynamic evolution of high-temperature heat waves,and the emergency and decision-making of heat wave disasters.The full text aims to establish a day-to-day change model of urban surface temperature,analyze the distribution of hourly surface temperature during the day,and further explore the dynamic evolution characteristics of urban heat islands.The research area selected in this thesis is mainly part of Beijing,and the data sources are Landsat 8,FY-4 temperature products and other remote sensing images on September 2,2019.The research contents of the full text mainly includes the following four parts: First,based on the 6S radiative transfer model,hourly direct solar radiation energy and scattered radiant energy from 8 to 17 hours are simulated;then,based on the surface radiation transfer model,the Landsat 8 red band and Downward radiation in the near-infrared band as well as the the surface reflectance and NDVI at the Landsat 8 transit time are took into consideration;thirdly,combined with a geographically weighted regression model,this article simulates the surface temperature with a temporal resolution of 1 hour and a spatial resolution of 1 km;finally,the simulated surface Temperature in this thesis is used to analyze the dynamic evolution of urban heat islands.The detailed research contents and main conclusions are as follows:1.The Combination of 6S radiative transmission model and surface radiative transmission model are used to simulate the downward radiation of Landsat 8 red and near-infrared bands with a spatiotemporal resolution of 1h / 30 m.And then the article calculated the approximate short-wave downward radiation Edown at each time as well as the Surface reflectance and NDVI at the time of wave crossing.The research results demonstrate that the average downward radiation in the red and near-infrared bands from 8:00 to 17:00 show an upward trend and then a downward trend;the average downward radiation received at the surface at noon is relatively large,and the average downward radiation received at 8 and 17 hours is small.Its maximum value is at about 12 o'clock.This thesis simulates the NDVI at the transit time.Because the NDVI change range within a day is small,it is assumed that the NDVI at the same time is the same as the scaling factor of the hourly downscaling model of surface temperature during the day.2.The article constructed downscaling model of land surface temperature during the day and time and analyzed urban heat island dynamic evolution.Firstly,the simulated Landsat 8 transit time NDVI,the time-lapse short-wave downlink radiation approximation Edown,and the DEM data are used as scale factors,and the Fengyun No.4 surface temperature product is used to establish a geographically weighted regression model to obtain hourly regression coefficients and residuals.The kriging interpolation method was used to interpolates the regression coefficients and residuals to 1KM,and then scales the simulated NDVI,short-wave downlink radiation approximation,and DEM data to a spatial resolution of 1KM by means of spatial aggregation.Regression coefficients,residuals,and scale factors at 1KM spatial resolution were used to calculate the hourly surface temperature.Based on the simulated daytime hourly surface temperature,the dynamic evolution of urban heat islands was analyzed.The research results show that due to different surface properties and heat absorption,the surface temperature distribution of cities and suburbs is significantly different,that is,the city temperature is higher than the suburban temperature.The densely populated and less green areas have higher levels of heat island distribution and areas covered with vegetation has a lower-level heat island trend.From 08:00 to 17:00,the area of heat-free islands showed a downward trend as a whole,the proportion of weak heat islands and strong heat islands changed little,and the changes of heat islands of other levels were not obvious.