论文部分内容阅读
冠层鲜生物量是指示作物长势状况的重要指标,可用于诊断作物氮素营养状况。但常规测定生物量的方法需要破坏性取样,不利于帮助精准施肥管理决策。遥感技术作为快速分析手段可用于估测作物多种生理、生化参数,特别是借助植被指数。研究表明归一化差值植被指数(normalized difference vegetationindex,NDVI)、最优土壤调节指数(optimization of soil-adjusted vegetation index,OSAVI)等现已构建植被指数可用于准确估测中低生物量信息,但在较高冠层生物量条件下,它们的估测能力显著下降。文章旨在提出更通用的新植被指数用于估测冠层生物量,以提高遥感技术在较高生物量条件下的预测准确度。试验在加拿大魁北克省的St-Jean-sur-Richelieu地区进行,5年(2004~2008)试验中采集了玉米及小麦不同生育期的地面高光谱数据及对应生物量信息,并在2005年获取航空机载高光谱影像一景(compact airborne spectro-graphic imager,CASI)。研究提出了红边三角植被指数(red-edge triangular vegetation index,RTVI),并将其与一些常见植被指数进行比较分析。结果表明RTVI是最好的估测冠层生物量的指数,在较高生物量条件下仍保持其对生物量变化的高敏感性,其与生物量间的决定系数为0.96。基于CASI影像,RTVI预测生物量值与实际测定值间的决定系数(R2)为0.58,预测标准误差为0.44 kg.m-2,取得了较好的结果。
Canopy fresh biomass is an important indicator of crop status, which can be used to diagnose crop nitrogen status. However, conventional methods of determining biomass require destructive sampling, which is not conducive to helping fertilization management decisions. Remote sensing can be used as a fast analytical tool to estimate the physiological and biochemical parameters of crops, especially with the help of vegetation index. The results show that the NDVI and the optimal soil-adjusted vegetation index (OSAVI) can be used to accurately estimate the low and middle biomass information, However, their ability to estimate significantly decreased at higher canopy biomass. The aim of the paper is to propose a more general new vegetation index for estimating canopy biomass in order to improve the prediction accuracy of remote sensing techniques at higher biomass. The experiment was conducted in the St-Jean-sur-Richelieu region of Quebec, Canada. During the five years (2004-2008), the ground-based hyperspectral data and corresponding biomass information of maize and wheat at different growth stages were collected and obtained in 2005 Compact airborne spectro-graphic imager (CASI). A red-edge triangular vegetation index (RTVI) was proposed and compared with some common vegetation indices. The results showed that RTVI was the best index to estimate canopy biomass and maintained its high sensitivity to biomass change at higher biomass with a coefficient of determination of 0.96 for biomass. Based on the CASI images, the determination coefficient (R2) between the predicted biomass value of RTVI and the actual measured value was 0.58, and the standard error of prediction was 0.44 kg.m-2, and good results were obtained.