地块尺度冬油菜湿渍害遥感监测方法研究

发布时间：2018-04-21 15:19

本文选题：地块 + 尺度　；参考：《浙江大学》2017年硕士论文

【摘要】：油菜是长江中下游地区的主要越冬作物之一,在我国粮油生产中占有重要地位。但由于该地区春季湿润多雨,导致油菜在春季经常遭受大面积的湿渍害,造成严重的产量损失。以往人们获取数据大多通过实验室分析以及田间调查、测量的方法。然而这些方法收集数据时效性较差,收集大尺度数据比较困难。近年来,随着卫星通信技术、空间定位技术、遥感技术和地理信息系统等对地观测技术的迅速发展和地球环境变化的加剧,人们对卫星遥感数据质量和数量的要求在不断提高。在农业遥感应用领域,用于农作物生长情况的监测与分析、病虫害的预测以及农作物的估产更是对高时间、高空间分辨率遥感影像的使用提出了极高的要求。本研究基于高分辨率多源多时相遥感卫星数据,以试验数据为基础,准确获取了湿渍害胁迫下油菜的生长信息,利用多种回归模型估算了湿渍害胁迫下的油菜生长状况,并制作油菜长势监测动态图,最后,对湿渍害胁迫下油菜的生长进行一定的定性分析和定量研究。主要研究内容和结论如下:(1)通过开展盆栽控制试验研究了油菜苗期和开花期湿渍害胁迫对生物量、叶面积和产量的影响。结果表明:在油菜苗期、开花期内受湿渍害胁迫,都会对油菜生长发育和产量产生影响,且开花期油菜对湿渍害的反应比苗期更为敏感。在油菜苗期受湿渍害水分胁迫超过15天、开花期超过10天时,会对油菜的生长发育产生显著性影响。因而在油菜生长发育过程中,尤其是在开花期,应该及时进行排渍工作,从而减轻湿渍害水分胁迫对油菜生长的影响。(2)开展田间试验,分别在油菜苗期、开花期进行20天的不同湿渍害水分胁迫处理,同时进行星地同步观测,将田间测量的LAI与从高分辨率卫星遥感数据计算得到的植被指数进行指数建模,来模拟油菜开花期前的LAI生长变化,检验表明GNDVI-LAI指数回归模型的建模效果最好,其验证精度最高。之后用GNDVI-LAI指数回归模型绘制油菜季节性长势监测动态图。(3)在油菜的苗期、开花期、角果期三个时期分别进行不同湿渍害处理的田间小区试验,同时进行星地同步观测。同时建立了五种回归模型,并使用最优回归模型来估算冬油菜的AGB,结果表明,NDVI和AGB建立的幂函数回归模型结果最好,其中建模R~2为0.75,验证指标RMSE为100.45 g/m2,rRMSE为20%。然后绘制油菜AGB季节性生长图,研究分析湿渍害胁迫下的油菜AGB的变异性,并对湿渍害胁迫下油菜的生长进行监测,说明高分辨率遥感数据能很好地反应出不同湿渍害胁迫处理在油菜不同生育期的生长差异,且结果与本研究中第二章盆栽控制试验结果较为一致,湿渍害胁迫对油菜生长的影响开花期大于苗期,淹水处理大于渍水处理。
[Abstract]:Rape is one of the main winter crops in the middle and lower reaches of the Yangtze River and plays an important role in the production of grain and oil in China. However, because of the wet and rainy spring in this area, rapeseed is often subjected to a large area of wet damage in spring, resulting in a serious loss of yield. In the past, most of the data were obtained by laboratory analysis and field survey. However, these methods are not effective in collecting data, and it is difficult to collect large scale data. In recent years, with the rapid development of earth observation technology, such as satellite communication technology, space positioning technology, remote sensing technology and geographic information system, and the aggravation of the Earth environment, the quality and quantity of satellite remote sensing data have been increasing. In the field of agricultural remote sensing application, the application of crop growth monitoring and analysis, the prediction of pests and diseases and the estimation of crop yield have put forward very high requirements for the use of high-time and high-spatial resolution remote sensing images. Based on high-resolution multi-source and multi-temporal remote sensing satellite data and experimental data, the growth information of rapeseed under wet waterlogging stress was obtained accurately, and the growth status of rape under wet waterlogging stress was estimated by multiple regression models. The dynamic map of rapeseed growth monitoring was made. Finally, the growth of rape under wet waterlogging stress was analyzed qualitatively and quantitatively. The main contents and conclusions are as follows: (1) the effects of wet waterlogging stress on biomass, leaf area and yield of rapeseed at seedling and flowering stage were studied by pot experiment. The results showed that the growth and yield of rapeseed were affected by the stress of wet waterlogging in flowering stage and seedling stage, and the response of rape in flowering stage was more sensitive than that in seedling stage. The growth and development of rapeseed were significantly affected by moisture stress for more than 15 days and flowering for more than 10 days at seedling stage. Therefore, in the process of rape growth and development, especially in the flowering period, it is necessary to carry out the soaking work in time, so as to reduce the effect of moisture stress on the growth of rapeseed, and carry out field experiments, respectively, at the seedling stage of rapeseed. After 20 days of water stress treatment with different wetting damage in flowering period, the LAI measured in the field and vegetation index calculated from high resolution satellite remote sensing data were modeled exponentially by simultaneous observation of satellite and ground. To simulate the change of LAI growth before flowering period of rapeseed, the test shows that the modeling effect of GNDVI-LAI index regression model is the best, and the accuracy of verification is the highest. Then the dynamic map of seasonal growth monitoring of rape was plotted by GNDVI-LAI index regression model. The plot experiments were carried out in three stages of rape seedling, flowering and pod respectively, and the field experiments were carried out simultaneously. At the same time, satellite and ground synchronous observation was carried out. At the same time, five regression models were established, and the optimal regression model was used to estimate the AGBs of winter rapeseed. The results showed that the power function regression model established by AGB and AGB was the best, in which the model RN-2 was 0.75, and the verification index RMSE was 100.45 g / m2 rRMSE was 20g / m ~ (-1). Then the seasonal growth map of rape AGB was drawn, and the variation of AGB in rape under wet waterlogging stress was analyzed, and the growth of rape under wet waterlogging stress was monitored. The results showed that the high resolution remote sensing data could well reflect the growth difference of rape under different wet waterlogging stress, and the results were consistent with the results of pot experiment in the second chapter of this study. The effect of wet waterlogging stress on rape growth was greater in flowering stage than in seedling stage and in flooding treatment than in waterlogging treatment.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S565.4;S127;S422

【相似文献】