基于DART模型的盐地碱蓬植被指数饱和问题分析

植被指数饱和问题直接影响植被参数的遥感估算精度.基于DART模型定量分析盐地碱蓬群落植被指数对LAI的响应，讨论不同植被指数的饱和问题，研究发现：1) 盐地碱蓬LAI与株数之间呈显著线性相关，R2为0.99，RMSE为0.008.2) TSAVI抗饱和性最好，NDVI及RVI次之，EVI、SAVI一般，PVI、MVI及MSAVI最差.3) RVI、NDVI及TSAVI敏感度最优，且LAI〈0.68时，SRVI〉SNDVI〉STSAVI；0.68≤LAI〈2时，SNDVI〉SRVI〉STSAVI；LAI≥2时，TSAVI最优，RVI和NDVI出现饱和现象；EVI、SAVI对LAI敏感性一般；剩余3种植被指数对LAI的敏感性弱.对于辽河口滨海湿地典型植被盐地碱蓬，结合实测LAI数据，初步分析认为TSAVI、NDVI及RVI可以作为较强适用性植被指数来进行相关研究，能够区分不同LAI下的盐地碱蓬群落.

Analysis on saturation of Suaeda salsa vegetation index based on DART model

Vegetation index (VI) saturation directly affects the accuracy of remote sensing estimation of vegetation parameters.Based on DART model， the response of Suaeda salsa community vegetation index to LAI was quantitatively analyzed， and the saturation of different vegetation indices was discussed. The results are shown as follows. 1) There was a significant linear correlation between LAI and plant number with R2 of 0.99 and RMSE of 0.008. 2) TSAVI had the best anti-saturation ability， followed by NDVI and RVI.The anti-saturation ablilities of EVI and SAVI are moderate， and those of PVI， MVI and MSAVI are the worst. 3) When LAI〈0.68， SRVI〉SNDVI〉STSAVI； when 0.68≤LAI〈2， SNDVI〉SRVI 〉STSAVI； and when LAI ≥ 2， TSAVI is the best， RVI and NDVI are saturated. EVI and SAVI are generally sensitive to LAI and the sensitivity of the remaining three vegetation indexes to LAI is weak.For Suaeda salsa， a typical vegetation in the Liaohekou Coastal Wetland， preliminary analysis combined with the measured LAI suggests that TSAVI， NDVI and RVI have good accuracy in estimating Suaeda salsa LAI， and are able to distinguish Suaeda salsa communities under different LAI.