浒苔藻体时间序列光谱特征的曲线估计  

CURVE ESTIMATION BASED ON TIME SERIES REFLECTANCE SPECTRA OF ENTERMORPHA PROLIFERA

作　　者：卢霞[1] 谢宏全[1] 周立[1] 王春艳[1] 

机构地区：[1]淮海工学院测绘工程学院,江苏连云港222005

出　　处：《海洋湖沼通报》2012年第3期96-102,共7页Transactions of Oceanology and Limnology

基　　金：江苏省自然科学基金"浒苔全生命周期光谱特征研究"(BK2009164);江苏省教育厅项目"典型滨海盐土植被高光谱检测"(10KJB170001)资助

摘　　要：选用相同自然光照和固定水环境条件下的3种密度梯度(0.5,2.5和5.0g/L)的浒苔藻体反射光谱为研究对象,通过单变量参数定义提取了高光谱特征参量,并选用能较好体现浒苔藻体光谱特征差异的特征变量如黄边位置、蓝边位置、归一化植被指数NDVI,比值植被指数RVI[719,680]和VI[555,680],应用曲线模拟法对这些具有时间序列特点的高光谱特征变量进行曲线估计,并分析其演变规律。结果表明:黄边位置无法模拟。应用二次函数或三次函数能高精度地模拟各个特征变量,蓝边位置曲线估计模型的判定系数最高达0.498;归一化植被指数NDVI为0.926,比值植被指数RVI[719,680]为0.729,比值植被指数VI[555,680]为0.765。这些特征变量随着时间的推移而呈现先上升后下降或先下降后上升再下降的起伏规律,演变趋势与浒苔藻体密度存在非线性关系;除了蓝边位置和黄边位置外,其他特征参量均可较好地描述浒苔藻体反射光谱特征的演变趋势。该研究为深入进行浒苔藻体的定量遥感研究提供数据支持和技术支撑。Entermorpha prolifera samples were collected from Lianyungang sea area, whose water environment was not changed and natural illumination condition was controlled. Three different densities of 0.5g/L, 2.5g/L, and 5.0g/L were designed. Reflectance spectra dur ing the life cycle of entermorpha prolifera were measured by using ASD Field HandHeld spectrometer. Based on the reflectance spectra of Entermorpha prolifera, some hyperspectral single variables such as yellow edge position, blue edge position, NDVI [719,6801, RVI [719,6801, and RVI[555,680] were extracted and chosen because they can embody better the difference of reflectance spectra of entermorpha prolifera. Then, curve estimation models were established by using curve estimation method for these five time series spectra character variables. The results indicated that yellow edge position cannot be effectively estimated, while other variables can be estimated by using quadratic or cubic functions. The highest Rz for blue edge position, NDVI, RVI[719,6801, and RVI[555,6801 were 0. 498, 0. 926, 0. 729, and 0. 765 respectively. These variables appeared fluctuating rule of creasing or de- creasing with the lapse of time. Therefore, the change tendency of these variables existed non-linear relation with the density of entermorpha prolifera. Other parameters such as REP, NDVI, RVI[719,680], and RVI[555,680] can be used to describe the development rule except blue edge position and yellow edge position. The research would provide data and technique support for further quantitative remote sensing of entermorpha prolifera.

关 键 词：浒苔 反射光谱 单变量 时间序列 曲线估计 

分 类 号：P237[天文地球—摄影测量与遥感]