Multi-scale Dynamic Monitoring Of Urban Surface Thermal Environment

With the continuous advancement of urbanization,a series of urban ecological and environmental problems have arisen.The key to solve the urban ecological and environmental problems is to carry out urban thermal environment dynamic observation.In recent years,with the continuous development of satellite remote sensing technology,it is possible to realize large-scale observation of urban space thermal environment by remote sensing technology.Conventional satellite sensors,such as ASTER,MODIS and Landsat,capture surface temperatures with a spatial resolution of 30 to 1 km.However,this spatial resolution cannot reflect the thermal characteristics of urban space at fine scale.In addition,the revisit period of such satellite sensors is long,which leads to the low temporal resolution of remote sensing image data.Therefore,these methods are not sufficient to assess urban LST variability throughout the day or to capture spatial heterogeneity of urban LST at a small scale.The UAV has the characteristics of on-demand flight and high spatial resolution,which is a new technology that can cope with this challenge.The research results show that on a sunny summer daytime observation day,the goodness of fit R~2 between the temperature value observed by the drone and the ground in-situ measurement value is R~2=0.73 during the morning to noon time period,and R~2=0.83 during the noon time period.R~2=0.85 for the time period from noon to night,and each time period showed a high degree of goodness of fit.On a clear summer night,the goodness of fit between the temperature value observed by the drone and the ground in-situ measurement value is R~2=0.45.Except for 3:16AM and 5:16AM,the goodness of fit is shown in other time periods.When the UAV is used for observation at night,the accuracy is obviously reduced.During the summer rainy days,two sunny rain events were observed.The goodness of fit between the temperature value observed by the drone in the first rainfall event and the ground in-situ measurement value are R~2=0.94 and R~2=0.91,respectively.The rainfall weather model is the best among all observations.The R~2 of the second rainfall event was R~2=0.067 and R~2=0.76,respectively.It is concluded that when drones are used for observational research before rainfall,they may be affected by the influx of clouds,resulting in inaccurate observation results.In addition,the UAV carried a thermal infrared lens to continuously observe and study the surface temperature of different underlying surfaces under different weather conditions.The results of the study show that drones overestimate the temperature of some types of land cover during the period from morning to noon on a sunny day.Specifically:the temperature of the cement floor was overestimated by the drone by 1°C,the temperature of the asphalt floor only differed by 0.04°C,and the surface water temperature of the No.1 lake and the No.3 lake were overestimated by the drone by 1.3?and 2?respectively.the temperature of Lake No.2 is almost the same,and the temperature of the underlying surface of the grass is overestimated by the drone by 3.6?.The difference between UAV observation data and ground measurement data during the noon period is very small,which is why many studies using UAVs to observe surface temperature choose to observe at noon.From noon to afternoon,the influx of clouds will cause inaccurate drone observations,and drones often underestimate the surface temperature.At noon,the temperature difference between the two observation methods is the largest(12.5°C).However,at night the temperature difference of the surface under the grass is within 2?.At night,both the standard deviation of the observation data and the temperature difference between the two observation methods are small(0.1-2?).Observational research on rainfall events on sunny days found that after rainfall on a sunny day,the thermal runoff formed by the rainfall quickly reached a state of thermal equilibrium with the underlying surface of the hard surface about 10 minutes after the rainfall.Before rainfall,the lake water absorbs heat radiation,and the water temperature is relatively high;when the rainfall begins,the temperature of the rainwater is lower than the temperature of the lake,which has a cooling effect on the lake;in the middle of the rainfall,runoff is formed,and the temperature of the runoff is higher than the surface water temperature of the lake,so the lake temperature rises.The surface thermal runoff in the study area does raise the temperature of the lake surface.