基于MNDWI水体指数的ASTER与ETM＋影像交互比较研究

Cross Comparison of Modified Normalized Difference Water Index(MNDWI) between ASTER and ETM+ Sensor Imagery

世纪之交以来相继发射的各种新型传感器给遥感领域带来了一个新的热门课题,即多种传感器影像之间的交互比较（Cross Comparison）。通过3对同日过空的ASTER与ETM＋影像对,分别基于影像原始DN值、ICM模型反射率和IACM模型反射率数据来对这两种传感器影像的水体指数（MNDWI）进行综合对比,并提供了两种传感器MNDWI指数互为转换的关系方程。结果表明,ASTER与ETM＋影像分别基于三类数据（DN、ICM、IACM）计算的MNDWI指数间,均存在很强的线性相关关系,决定系数都大于或接近0.9。精度验证表明,ASTER与ETM＋影像MNDWI指数互为转换的关系方程具有很高的精度,其误差RMSE介于0.057-0.073之间;且以基于IACM模型反射率的结果最为稳定、精度最高。因此,推荐使用该模型数据来进行这两种传感器影像MNDWI指数的互为转换。

The successful launches of many new advanced sensors on board remote sensing satellites since the latest 1990s have brought a new remote sensing study field,i.e.,cross comparison between multi-sensor imagery.This paper focuses on the comparison between Landsat-7 ETM＋ and ASTER sensor imagery because the cross comparison between the two sensors has not yet been implemented.This comparison was based on the Modified Normalized Difference Water Index（MNDWI）,derived from three synchronous image pairs of two sensors,and was carried out using three different data types,which are digital number（DN）-based,illumination correction model （ICM）-based and illumination and atmospheric correction model（IACM）-based data.The regression analysis for the MNDWI images of the two sensors has revealed the interconversion relationship between the MNDWI images of the two sensors,and thus the cross-sensor MNDWI conversion equations for the two sensors.The analysis also indicates that the MNDWI of ASTER imagery has a strong positive correlation with that of ETM＋ imagery with R2 values of greater than or close to 0.9.The accuracy verification shows that all the conversion equations have high accuracy with a RMSE ranging from 0.04 to 0.09,and the IACM-based equation has achieved the highest accuracy among the three data types.Therefore,this paper suggests the use of IACM-corrected data for the interconversion of MNDWI between the two sensors.