摘要
准确、及时地获取大面积地表土壤水分信息在农业科学领域具有较大的应用潜力。本文以地域特色突出的广东省徐闻县为试验区,联合使用双极化ENVISAT-ASAR数据和光学遥感MODIS数据,充分利用MODIS数据的高重复覆盖率,提取试验日期内的归一化差分水指数(NDWI),计算各试验点所对应的植被含水量(VWC)数据,相较于NDVI,NDWI具有不易饱和的特点,更加适合于本次研究所在的高植被覆盖区。最后利用"水—云模型"从ASAR数据总的后向散射中去除植被的影响,建立裸土后向散射系数与实测土壤含水量之间的关系,拟合结果VV极化优于HH极化,相关系数为VV极化R=0.865,HH极化R=0.676,总体上针对农作物覆盖地表土壤水分变化的估算算法还需要更进一步发展和改进,以提高反演精度。
It has been greatly applied in agricultural sciences to retrieve large area soil moisture information accurately and promptly. The study area is located in Xuwen County which has regional features prominently. Dual-polarization ENVISAT-ASAR and MODIS images were used to retrieve soil moisture. The estimation of vegetation water content （VWC） of experiment regions was performed using MODIS-derived normalized difference water index （NDWI） near the experiment date. Compared with NDVI, NDWI is more suitable to high vegetation covering in the study area. NDWI can reflect changes when NDVI is saturated along with cropgrowth. In the end the vegetation effects were eliminated from the total ENVISAT-ASAR backscattering coefficients by using ＂water-cloud＂ model, and then, we simulated the relation between the backscattering coefficient of bare soil and ground measured soil moisture after eliminating vegetation effect. The results showed that VV polarization fitting is better than HH polarization, with correlation coefficient 0. 865 in VV polarization and 0. 676 in HH polarization. The estimation of soil moisture content in vegetated areas should be further improved in the future.
出处
《干旱地区农业研究》
CSCD
北大核心
2008年第3期39-43,共5页
Agricultural Research in the Arid Areas
