论文部分内容阅读
分别利用6S模型和FLAASH模型对TM影像进行了大气校正,建立了两种校正模型下的叶面积指数(LAI)与多种植被指数(VI)间的线性与非线性统计回归关系,并从植被指数角度和大气校正模型角度对LAI-VI回归相关性进行了分析,最后通过验证数据组LAI预测值(Y)与LAI实测值(X)的RMSE均方根误差及Nash-Sutcliffe效率系数对各模型下的LAI反演结果进行了精度对比验证。结果表明:除RVI外,两种大气校正模型下的LAI与SAVI、MSAVI有良好的相关性,其中6S模型下的LAI-MSAVI一元二次多项式回归模型拟合优度最佳;不同大气校正模型对LAI-VI回归方程的相关性有显著影响,6S模型的LAI反演精度优于FLAASH模型;借助遥感技术定量提取植被生理参数时,应慎重选择适宜的大气校正模型。
The 6S model and the FLAASH model were used to calibrate the TM images respectively. The linear and nonlinear statistical regression relationships between leaf area index (LAI) and multiple vegetation indices (VI) under the two calibration models were established. The relationship between the LAI-VI regression and the LAI-VI regression was analyzed from the angle of the index and the atmospheric correction model. Finally, the RMSE root mean square error (RMSE) and the Nash-Sutcliffe efficiency coefficient of the LAI predicted value (Y) LAI inversion results under the model were verified by the accuracy comparison. The results show that LAI has good correlation with SAVI and MSAVI except for RVI, and the LAI-MSAVI quadratic polynomial regression model under 6S model has the best goodness of fit. For different atmospheric correction models LAI-VI regression equation has a significant impact on the correlation, 6S model LAI inversion accuracy is better than the FLAASH model; quantitative extraction of vegetation physiological parameters by remote sensing, should be carefully selected appropriate atmospheric correction model.