【作者】 于亚娇； 龚龑； 方圣辉； 彭漪； 杨凯丽； 袁宁鸽； 吴贤婷； 朱仁山；

【Author】 Yajiao Yu;Yan Gong;Shenghui Fang;Yi Peng;Kaili Yang;Ningge Yuan;Xianting Wu;Renshan Zhu;School of Remote Sensing and Information Engineering, Wuhan University;Lab for Remote Sensing of Crop Phenotyping, Wuhan University;College of Life Sciences, Wuhan University;

【机构】 武汉大学遥感信息工程学院； 武汉大学遥感农作物表型实验室； 武汉大学生命科学学院；

【摘要】 利用遥感技术准确估算水稻产量对农业决策至关重要。经验统计模型作为最常用的遥感技术估产方法,只关注到水稻在某个或多个时刻的生长状况,忽略了水稻产量的形成是个动态变化的过程这一特点。本文以2018年42种杂交水稻的冠层反射率为基础数据,计算并筛选出最佳的光谱累积量信息,将其与水稻产量进行回归。结果表明,ACC(72,98)在本文所选的9个植被指数中都与水稻产量呈现出较好的相关性,优于孕穗期的估产精度,验证了本文估产方法的可行性。分蘖后期至孕穗期的光谱累积量作为多生育期植被信息,可表征水稻在关键生育期的光合作用能力总和。因其包含多个生育期的植被信息而优于单生育期估产精度,又因其不包含杂糅无效的植被信息而优于全生育期估产精度。更多还原

【Abstract】 Accurate estimation of rice yield using remote sensing technology is essential for agricultural decisionmaking. As the most common method for estimating rice yield by remote sensing technology, empirical statistical model only focuses on the limited growth status of rice. However, it ignores the fact that the formation of rice yield is a dynamic process. In this paper, based on the canopy reflectivity data of 42 hybrid rice species in 2018, the best information of spectral accumulation is calculated and screened, and it is regressed with the yield. The results show that ACC(72,98) has a good correlation with rice yield among the nine vegetation indices selected in this paper. Additionally, ACC(72,98) is better than the estimated yield accuracy of the booting stage, which verifies the effectiveness of the rice yield estimation method proposed in this paper. Moreover, spectral accumulation from the late tillering stage to the booting stage, as multifertility vegetation information, can characterize the total photosynthesis ability of rice in key growth periods. As a multi-stage method which contains more vegetation information,it achieves better yield estimation accuracy than the singlestage method. Otherwise, it is better than the estimation accuracy of the entire growth stage, because ineffective vegetation information is excluded by this indicator.更多还原