您的位置:首页  > 论文页面

基于太阳辐射能量的地表温度日间逐小时变化模拟研究

发表时间:2020-10-10  浏览量:973  下载量:139
全部作者: 邓帆,陶于祥,罗小波
作者单位: 重庆邮电大学计算机科学与技术学院;重庆市气象科学研究所
摘 要: 现有的卫星热红外数据在时间与空间分辨率上存在矛盾,导致难以获取高时空分辨率地表温度(land surface temperature,LST),制约了城市热环境精细化监测能力。将辐射能量模型与地理加权回归(geographical weighted regression,GWR)模型集成用于地表温度模拟,基于6S(Second Simulation of Satellite Signal in the Solar Spectrum)辐射传输模型模拟不同太阳位置时地表像元接收的直射辐射能量与散射辐射能量,利用美国陆地卫星影像(Landsat 8 TM)数据结合地表辐射传输方程模拟不同时刻的波段下行辐射、短波净辐射近似值以及过境时刻的波段地表反射率,进一步模拟过境时刻的归一化植被指数(normalized differential vegetation index,NDVI)尺度因子。最后采用高时间分辨率的风云四号(FY-4)静止卫星数据,利用GWR降尺度方法模拟日间逐小时地表温度。为评估模拟地表温度的精度,将模拟结果与中分辨率成像光谱仪(moderate-resolution imaging spectroradiometer,MODIS)获取的相近时刻的地表温度产品数据进行对比验证。结果表明,模型模拟的精度均方根误差(root mean square error,RMSE)为1.37℃,平均绝对误差(mean absolute error,MAE)为1.04℃,R2为0.696 8,相对误差直方图显示大部分像元的模拟精度误差都小于3℃,表明能够较好地模拟地表温度。
关 键 词: 地球科学其他学科;辐射能量模拟;6S;归一化植被指数;逐时地表温度
Title: Simulation study on hourly change of land surface temperature within the day based on solar radiant energy
Author: DENG Fan, TAO Yuxiang, LUO Xiaobo
Organization: School of Computer Science and Technology, Chongqing University of Posts and Telecommunications; Chongqing Institute of Meteorological Science
Abstract: The existing satellite thermal infrared data has a contradiction in temporal and spatial resolution, which makes it difficult to obtain the land surface temperature (LST) of high spatial and temporal resolution, which restricts the ability of refined monitoring of urban thermal environment. The radiant energy model and the geographic weighted regression (GWR) model were integrated for land surface temperature simulation. Based on the 6S (Second Simulation of Satellite Signal in the Solar Spectrum) radiative transmission model, the direct radiant energy and scattered radiant energy received by the surface pixels at different solar positions were simulated. The down-band radiation and the approximate value of short-wave net radiation at different times, and the surface reflectance of the band at transit time were simulated by combining the United States Landsat 8 TM data with the surface radiation transfer equation. Furthermore, the normalized differential vegetation index (NDVI) scale factor at transit time was simulated. Finally, the simulation of hourly land surface temperature during the day was carried out using highly time-resolved FY-4 geostationary satellite data by GWR downscaling method. In order to evaluate the accuracy of the simulated land surface temperature, the simulation results were verified with the land surface temperature product data obtained at a similar time from the medium-resolution imaging spectrometer (MODIS). The results show that the root mean square error (RMSE) of the model simulation is 1.37℃, the mean absolute error (MAE) is 1.04℃, and the R2 is 0.696 8. The histogram of relative error indicates that the simulation accuracy errors of most pixels are less than 3℃, which can better simulate the land surface temperature.
Key words: other subjects of earth science; radiant energy simulation; 6S; normalized differential vegetation index; hourly land surface temperature
发表期数: 2020年9月第3期
引用格式: 邓帆,陶于祥,罗小波. 基于太阳辐射能量的地表温度日间逐小时变化模拟研究[J]. 中国科技论文在线精品论文,2020,13(3):370-381.
 
8 评论数 0
暂无评论
友情链接