中国空间科学技术

中国空间科学技术 ›› 2022, Vol. 42 ›› Issue (3): 93-104.doi: 10.16708/j.cnki.1000-758X.2022.0040

• 论文 • 上一篇    下一篇

基于DTW的HJ-1B卫星地表温度产品质量检验

张琪，李大成，历华，段四波，张霄羽，韩启金，高海霞   

  1. 1太原理工大学，太原030024
    2中国科学院 空天信息创新研究院，北京100094
    3中国农业科学研究院，北京100081
    4山西大学，太原030006
    5中国资源卫星应用中心，北京100094
  • 出版日期:2022-06-25 发布日期:2022-06-22

Quality validation of HJ-1B satellite land surface temperature products based on DTW

ZHANG Qi,LI Dacheng,LI Hua,DUAN Sibo,ZHANG Xiaoyu,HAN Qijin,GAO Haixia   

  1. 1Taiyuan University of Technology,Taiyuan 030024,China
    2Aerospace Information Research Institute，Chinese Academy of Science,Beijing 100094,China
    3Chinese Academy of Agricultural Sciences,Beijing 100081,China
    4Shanxi University,Taiyuan 030006,China
    5China Centre for Resources Satellite Data and Application,Beijing 100094,China
  • Published:2022-06-25 Online:2022-06-22

PDF (PC)

63

摘要： HJ-1B卫星热红外数据应用广泛，但其地表温度反演产品的质量检验工作尚不完善。以黑河流域为研究区，利用普适性单通道算法得到HJ1B地表温度，基于7个地面站点（下垫面为荒漠、沙漠、植被、农作物、雪地和湿地）同步观测资料和MODIS地表温度产品（MOD11A1），引入动态时间规整方法（DTW）对站点处HJ1B地表温度进行验证。试验结果表明：HJ1B地表温度反演产品与地面观测值的偏差值在沙漠和荒漠站点为1K以内，均方根误差在05K左右；对于植被和农作物站点的偏差在2K以内，均方根误差为1~2K；基于DTW的验证对时序不匹配的数据评价结果与现有指标表现一致。HJ-1B地表温度反演产品与地面站点的相对偏差均低于其与MODIS地表温度反演产品的相对偏差。

关键词: HJ-1B卫星, 地表温度, 直接检验, 普适单通道算法, 黑河流域, 动态时间规整

Abstract: The HJ-1B satellite thermal infrared data is widely used,but the quality validation of its surface temperature retrieval products is not yet perfect.In response to this problem,part of multiple underlying surfaces in the Heihe River Basin was taken as the research area,HJ-1B data was used as the research data,and universal single-channel algorithm was used to invert the surface temperature of the different underlying surface research areas.The inversion results were verified using the ground observation site data of the Heihe River Basin and MODIS surface temperature data.Dynamic time warping was introduced to evaluate timing similarity.The results show that the relative deviation between the HJ-1B surface temperature inversion product and the ground observation value for desert and desert sites is within 1K,and the root mean square error(RMSE) is about 0.5K.The deviation for vegetation and crop sites is within 2K,and the RMSE is 1.2K.The evaluation result of the time-series mismatched data using DTW is consistent with the performance of existing indicators.The relative deviations between the HJ-1B surface temperature inversion products and the surface stations are lower than those between the HJ-1B surface temperature inversion products and the MODIS surface temperature inversion products.

Key words: HJ-1B satellite;land surface temperature, single-channel algorithm, direct validation；Heihe River Basin;DTW