基于多源遥感数据的冬小麦氮肥推荐的研究

发布时间：2018-11-25 12:30

【摘要】：小麦是我国重要的粮食作物,对小麦进行科学的田间管理是保证其优质高产的必需条件。氮素是在作物的生长过程中起到了关键作用的营养元素,适量的投入将会促进作物的光合作用,也将影响经济增长和发展,保证作物产量和品质。为确保产量,农民往往施入过多氮肥。大量氮肥的投入不但使氮肥利用率大大降低,给环境造成巨大危害,也显著降低了农民收益。准确探测作物氮素营养状况,在作物生长的关键时期以适宜养分量供给作物生长,提高氮肥利用率已成为当代农业研究的热点。近年来,随着遥感技术的飞速发展,其以访问频率高、适合大面积、无损作物等优势,已成为实现快速有效且无损监测冬小麦氮营养情况的一种新的技术手段和方法,它作为精细农业的重要组成部分,成为对现代农业的研究中最为重要的环节之一。归一化植被指数(NDVI)是一种能反映作物冠层结构的重要参数,其与作物的产量、群体光的分布以及光能的利用有着密切的关联。本文首先利用多年的冬小麦冠层光谱数据及产量数据,并与气象数据相结合,构建适用于我国北方的冬小麦潜在产量模型,在遥感影像中提取相应的NDVI值对模型进行验证。在田块尺度的施肥推荐方面,基于2013年覆盖国家精准农业研究示范基地的示范田的Geo Eye-1高分遥感数据,运用构建模型进行施肥推荐,得到施肥推荐处方图,推荐施肥量在合理范围内。在区域尺度施肥推荐方面,首先利用S-G滤波方法进行陕西省2010年至2014年MODIS NDVI时间序列数据的重建,根据陕西省典型地物特性曲线变化趋势,确定冬小麦与其他作物特性曲线的差异,提取了陕西省冬小麦的种植区域。其次,使用MODIS LAI数据产品MOD15A2多景数据,采用S-G上包络线滤波方法重构LAI时间序列,并对滤波后的数据进行Logistic曲线拟合,以曲线曲率的最大值作为冬小麦返青期开始的时间,得到冬小麦返青期监测结果。以监测到的返青期开始时间为基准,将所需累计积温转化为累计LST值,结合MODIS LST数据产品MOD11A2来监测拔节期。最后,以遥感推算的冬小麦物候期为依据对研究区进行区域划分,在划分的每一个区域内应用施肥推荐模型,得到施肥推荐处方图。研究结果表明,本文所构建模型适用于我国北方地区的施肥推荐,在田块与区域不同尺度上推荐的施肥量均在合理范围内,且能够有效的提高氮肥的利用效率,节约农业成本,且进一步说明利用遥感技术进行施肥推荐的可行性及具有一定的应用价值。
[Abstract]:Wheat is an important food crop in China. Scientific field management is a necessary condition to ensure high quality and high yield of wheat. Nitrogen plays a key role in the growth of crops. Proper input will promote photosynthesis, affect economic growth and development, and ensure crop yield and quality. To ensure yield, farmers often apply too much nitrogen fertilizer. The input of a large amount of nitrogen fertilizer not only greatly reduces the utilization efficiency of nitrogen fertilizer, causes great harm to the environment, but also significantly reduces the farmers' income. It has become a hot point in modern agricultural research to accurately detect the nitrogen nutrition status of crops, supply crop growth with suitable nutrient in the critical period of crop growth, and improve nitrogen fertilizer utilization efficiency. In recent years, with the rapid development of remote sensing technology, with the advantages of high access frequency, suitable for large area and no damage to crops, it has become a new technical means and method to realize fast, effective and non-destructive monitoring of nitrogen nutrition of winter wheat. As an important part of fine agriculture, it has become one of the most important links in the research of modern agriculture. Normalized vegetation index (NDVI) is an important parameter reflecting crop canopy structure, which is closely related to crop yield, distribution of population light and utilization of light energy. Based on the canopy spectral data and yield data of winter wheat for many years, and combined with meteorological data, a winter wheat potential yield model suitable for northern China was constructed in this paper. The corresponding NDVI values were extracted from remote sensing images to verify the model. In the field scale fertilization recommendation, based on the Geo Eye-1 high score remote sensing data of the demonstration field covering the national precision agriculture research demonstration base in 2013, using the model to carry on the fertilization recommendation, get the fertilizer recommendation prescription map. The recommended amount of fertilizer is within a reasonable range. In the area of regional scale fertilization recommendation, the S-G filtering method is firstly used to reconstruct the MODIS NDVI time series data from 2010 to 2014 in Shaanxi Province. The difference of characteristic curve between winter wheat and other crops was determined and the planting area of winter wheat in Shaanxi Province was extracted. Secondly, using MOD15A2 multi-scene data of MODIS LAI data product, S-G envelope filtering method is used to reconstruct LAI time series, and the filtered data is fitted by Logistic curve, and the maximum curvature of curve is taken as the beginning time of winter wheat returning to green. The monitoring results of winter wheat were obtained. Based on the monitored start time of the green stage, the accumulated temperature was converted into the cumulative LST value, and the jointing stage was monitored by the MODIS LST data product MOD11A2. Finally, according to the phenological period of winter wheat calculated by remote sensing, the study area was divided, and the fertilization recommendation model was applied to each area, and the recommended fertilizer prescription map was obtained. The results showed that the model was suitable for the recommendation of fertilization in the north of China. The recommended amount of fertilizer was within a reasonable range on different scales of fields and regions, and could effectively improve the utilization efficiency of nitrogen fertilizer and save the cost of agriculture. The feasibility and application value of using remote sensing technology to recommend fertilization are further explained.
【学位授予单位】：东北农业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：S512.11;S127

【相似文献】