基于ENVISat ASAR影像的湿地植被冠层下淹水范围提取以扎龙保护区为例

以扎龙自然保护区湿地为例,结合ENVISat ASAR多极化(HH/HV)雷达影像与传统的光学影像Landsat TM (band1~5,7),分析雷达影像后向散射系数与Landsat TM影像不同波段反射率在淹水植被、非淹水植被、明水面和裸土不同地表覆被类型的差异。选择训练样本,采用分类回归树(Classification and Regression Tree,CART)模型,分别对两种影像进行分类,可视化表达湿地植被淹水范围空间分布情况。基于实测的植被冠层下淹水范围与非淹水范围样本点对两种数据源的分类结果进行精度验证。结果表明:HH/HV极化影像中,植被覆盖下水体的后向散射系数与其他地表覆被类型有明显区别,分类结果总精度为79.49%,Kappa系数为0.70,湿地植被淹水范围提取精度较高。而TM影像分类结果中,由于部分地区植被覆盖水体,淹水植被分类误差较高。将雷达影像引入沼泽湿地研究,提高了植被淹水范围提取效果,为有效分析湿地生态水文过程提供基础,对湿地水资源合理利用及生物多样性保护具有重要意义。

Extraction of Flooding Extent under Wetland Vegetation Canopy Using ENVISat ASAR Imagery—A Case Study in the Zhalong Natural Reserve

Multi\|polarization (HH/HV) ENVISat ASAR imagery and conventional optical remotely sensed imagery of the Landsat TM (band1~5,7) were used in this study to investigate flooding extent of Zhalong Natural Reserve.Differences between backscattering coefficients of ASAR and different bands reflectance of Landsat TM imageries were conducted in different land cover types (flooded vegetation,non\|flooded vegetation,open water,bare soil).Moreover,the classification and regression tree (CART) algorithm was employed to extract spatial distribution of wetland vegetation flooding extent with two types of imageries based on selecting training samples while ground\|truth GPS samples were used to validate classification results of two data sources.The findings demonstrate that the backscattering coefficients of water under the vegetation canopy were different from other land cover types in HH/HV polarization imagery data.The overall accuracy was satisfactory (79.49%) with higher accuracy in extracting flooded vegetation;the kappa coefficient was 0.7.However,the results of flooded vegetation with TM imagery data appeared higher classification error in some regions where water was covered with vegetation.Consequently,the SAR imagery data plays a critical role in improving the accuracy of flooding extent extraction.Furthermore,the results reported here verify that it is effective to analysis the wetland eco-hydrological processes with SAR,which is very important for rational use of water resources of natural reserve and biodiversity conservation.