基于近地高光谱遥感的小麦叶片生长参数动态模型研究

Dynamic model for wheat leaf growth parameters based on in-situ hyperspectral remote sensing

【目的】探讨利用遥感技术实时监测小麦叶片生长动态变化。【方法】以生产中大面积推广应用的小麦品种周麦27为试验材料,在2个试验地点布置水氮耦合处理试验,依据相关回归分析筛选出对叶片氮含量(LNC)和叶面积指数(LAI)反应敏感的高光谱植被指数,进而确立了不同产量层次的植被指数生育进程动态模型。【结果】LNC和LAI与近红外短波段735~1075 nm呈显著正相关关系,而与可见光波段350~730 nm呈显著负相关关系。对LNC敏感的植被指数主要有AIVI、RES和mND924,而对LAI敏感的植被指数主要有ONLI、CI_(green)和MSR(800,670),以上2类植被指数和籽粒产量间关系均密切,表现较好的时期主要为拔节期至灌浆中期阶段。采用双LOGISTIC模拟模型方法,优选的方程能够较好地模拟植被指数的生育进程动态轨迹,模型精度(R^(2))随着产量水平的逐渐提高而增加,低产水平的精度相对较差(0.627~0.703),而高产及以上水平的R^(2)较高(0.868~0.972)。【结论】高光谱植被指数AIVI和CI_(green)用于评价不同产量水平下小麦叶片生长动态较为适宜,这为田间氮肥精确定量管理提供参考依据和技术支持。

【Objective】This study is designed to explore the application of remote sensing technology to monitor the dynamic change of wheat leaf growth.【Method】In this study,Zhoumai 27,a widely popularized and planted wheat cultivar,was used as the experimental material under water-nitrogen coupling test in different locations,and the hyperspectral vegetation indices sensitive to leaf nitrogen content(LNC)and leaf area index(LAI)were screened according to the correlation and regression analysis.And the dynamic growth process model of optimal vegetation index at different yield levels was established.【Result】The results showed that LNC and LAI were significantly positively correlated with 735-1075 nm in the near infrared short-wave band,but significantly negatively correlated with 350-730 nm in the visible band.The vegetation indices sensitive to LNC mainly included AIVI,RES and mND924,and the vegetation indices sensitive to LAI mainly were ONLI,CI_(green) and MSR(800,670).These two types of vegetation indexes were closely related to grain yield,and the better performance period was mainly from the jointing stage to the mid-filling stage.By using double Logistic simulation model,the optimal equation could better fit the dynamic path of growth process of vegetation index.The model accuracy(R^(2))increased with the increase of the output level,and R^(2) of low yield level was relatively poor(0.627-0.703),while that of high yield was higher(0.868 to 0.972).【Conclusion】AIVI and CI_(green) were more suitable for evaluating wheat leaf growth and dynamics,which would provide reference and technical support for accurate quantitative management of nitrogen fertilizer in wheat field.