论文部分内容阅读
针对当前基于被动微波遥感重建地表温度的统计方法难以实现大面积复杂下垫面区域数值重建的问题,提出了基于统计模型与滤波算法联合的地表温度重建方法。从时间序列角度探索地表温度与地表亮温的相关性,建立二者之间的统计模型,不需要进行地物分类,能有效避免下垫面复杂度对重建精度的影响;遍历像元,实现对大面积区域的数值重建。此外,采用滤波算法对基于统计模型的结果进行改正,利用地表温度时间序列的周期性进一步控制反演误差。针对MODIS地表温度产品重建的实验结果表明:所提算法精度明显提高,可用于各类下垫面覆盖区域的地表温度重建。
Aiming at the current problem that it is difficult to reconstruct the surface temperature based on passive microwave remote sensing, it is difficult to reconstruct large area of complex underlying surface. A new method of surface temperature reconstruction based on statistical model and filtering algorithm is proposed. From the perspective of time series, the correlation between surface temperature and surface brightness temperature is explored, and the statistical model between the two is established. The classification of features does not need to be done, which can effectively avoid the influence of underlying surface complexity on the reconstruction accuracy. Reconstruction of large area. In addition, the filtering algorithm is used to correct the results based on the statistical model, and the inversion error is further controlled by the periodicity of the surface temperature time series. Experimental results on MODIS surface temperature reconstruction show that the accuracy of the proposed method is obviously improved and can be used to reconstruct the surface temperature of various underlying surface covered areas.