基于近地遥感系统的小麦玉米冠层RVI和NDVI获取影响因素分析

以大型喷灌机为平台的近地遥感技术可有效观测作物的生长状态，对田间生产管理和作物水肥需求特性等研究具有十分重要的意义。由于在遥感观测过程中，作物冠层具有二向反射特性，因此不同观测方式会影响遥感观测结果。通过自行搭建的近地遥感系统模拟大型喷灌机平台的实地观测条件，使用双通道光谱传感器获取小麦与玉米冠层的光谱反射率信息，引入变异系数CV对由冠层二向反射特性引起的信息数据变幅进行量化，并采用影响因素权重W分析各观测参数对数据变幅的影响程度。通过获取2019年冬小麦返青期至灌浆期、夏玉米V7-V14生育期的冠层近红外波段（810 nm）和红光波段（650 nm）的反射率数据，分析多种观测因素对比值植被指数（RVI）数据和植被归一化指数（NDVI）数据的影响。结果表明，观测高度（0.5~2.5 m）、观测频率（2~60次·min-1）和移动速度（0~4 m·min-1）与观测结果无显著相关关系（p＞0.05），观测时刻（8:00-18:00）、观测天顶角（-60°~60°）和观测方位角（0°~180°）与观测结果相关关系极为显著（p＜0.01）；小麦和玉米的冠层RVI、NDVI数据获取结果主要取决于冠层覆盖程度，在相同叶面积指数（LAI）情况下观测结果也会因观测时刻、观测方位角和观测天顶角的差异而受到不同程度的影响；冠层光谱反射率信息二向反射特性明显，小麦冠层RVI和NDVI变异系数分别为15%~50%和2%~50%，玉米冠层RVI和NDVI变异系数分别为10%~33%和18%~39%；进行观测时，应尽量选择在太阳天顶角较稳定的12:00-14:00时段，并尽量缩短观测时长，还应选择固定的观测角度，注意阴影效应与热点效应的影响；此外，在小麦返青至拔节期、抽穗至扬花期获取RVI和NDVI时，还应分别注意观测天顶角、观测时刻对测量精度的干扰。研究结果可为快速获取高精度的小麦、玉米冠层光谱反射率数据提供技术支撑。

Analysis of Influencing Factors in Wheat/Maize Canopy RVI and NDVI Acquisition Using Ground-Based Remote Sensing System

Implement ground-based remote sensing technology in a large-scale sprinkler irrigation system can monitor crop growth status and has played an important role in detecting of crop water and nutrient demands and field management of crop production. Crop canopy has bidirectional reflectance characteristics, so different observation methods used in ground-based remote sensing will affect the accuracy of canopy spectral reflectance measurement. This study used a self-built ground-based remote sensing system to simulate the field observation conditions of large sprinkler irrigation machinery, obtained spectral reflectance information of wheat and maize canopy through multi-spectral optical sensors, and quantified the variation of the crop canopy information data caused by the canopy bidirectional reflection characteristics by the coefficient of variation CV, and analyzed the effect of the observation conditions on the crop canopy spectral reflectance measurements through influence factor weight W. Reflectance in the red band (650 nm) and reflectance in the near-infrared band (810 nm) of winter wheat at regreening to filling growth stages and summer maize at V7 to V14 growth stages were measured and recorded in 2019 season. Effect of various observation factors on the ratio of vegetation index (RVI) and the normalized index of vegetation (NDVI) was analyzed. Results showed that the correlation of observation height in 0.5~2.5 m, observation frequency in 2~60 min-1 and moving speed in 0~4 m·min-1 with the measurements of the canopy spectral reflectance characteristics were not significant (p>0.05); observation time in 8:00-18:00, observation zenith angle at -60°~60°, and observation azimuth angle from 0° to 180° had an extremely significant correlation with the measurements (p＜0.01). Results of canopy spectral reflectance measurement for wheat and maize depended mainly on the degree of the canopy coverage. Under the same leaf area index (LAI), the canopy spectral reflectance would also be affected by the observation time, observation azimuth and observation zenith angle: canopy spectral reflectance had significant bidirectional reflection characteristics. In wheat crop, the coefficients of RVI and NDVI variation were 15%~50% and 2%~50%，respectively, while in maize, they were 10%~33% and 18%~39%, respectively. When measuring RVI and NDVI with wheat and maize crops, the desired time for the measurement could be 12:00-14:00 because the solar zenith angle is relatively stable. The observation angle should be in a fixed angle, and also the influence of shadow effect and hot spot effect should be noticed. When measuring RVI and NDVI of wheat during regreening to jointing stage and heading to the flowering stage, close attention should be paid to the effects of observation zenith angle and observation time, respectively. This study performed a quantitative analysis of measuring canopy spectral reflectance with wheat and maize crops. The results obtained in the study could provide technical support for accurate and effective measurement of the crop canopy spectral reflectance.