【作者】 刘世峰；

【导师】 潘剑君；

【作者基本信息】 南京农业大学 ， 土壤学， 2010， 硕士

【摘要】 土壤是生活在地球上的人类赖以生存的物质基础和宝贵源泉。作为作物赖以生长并为全世界提供衣食的中介,土壤是植物生长繁育和生物生产的基地。而土壤中的各种元素含量是衡量土壤肥力,决定作物丰产与否的主要原因。因此,了解土壤肥力并根据作物生产的基本营养需求,进而认识到土壤对于作物的限制性因子对于农田的合理施肥和作物增收是至关重要的。由于农田土壤肥力空间分布存在差异,传统的均一施肥方式,容易导致局部施肥过多而造成农业面源污染,也同样会导致一些区域因施肥不足产生地力衰竭。另外,土壤中的各种元素含量是一个相对动态的变化过程。而进行配方施肥就必须准确掌握田间土的壤肥力状况,所以对土壤中的各种营养元素含量进行实时监测是必不可少的。传统的土壤肥力状况监测一般是实地采样,然后风干,磨土过筛,最后进行实验室化学分析。一方面投入大,浪费大量的人力物力；另一方面,监测土壤的周期也比较长,这就造成了土壤监测的时效性差。耕作者不能有效及时详尽地掌握田间土壤的肥力信息,难以做到及时有效的配方施肥去消除土壤对于作物的限制性因子。所以,寻找一种快速、有效、低投入的土壤肥力监测方法显得迫在眉睫。遥感技术是20世纪60年代兴起并迅速发展起来的一门综合性探测技术,是一种通过量测从地表获取电磁辐射进而推断地表参数的过程。土壤作为地球表面的最常见的地物,其对特定波段的电磁波反射率非常敏感。它的一些光谱反射率和本身的一些理化参数之间存在着很大的相关性,这就是利用遥感监测土壤肥力的内在机理。然而,农田土壤经常为植被所覆盖,裸露的时间比较短,影响了遥感技术在农田土壤肥力监测方面的应用。不过,国内外学者在遥感监测作物长势、营养状况方面取得了巨大成果。由于不同的土壤理化性质和施肥量会在一定程度上引起作物长势的差异,并最终反映在作物冠层光谱反射率上。这种机理的存在使得利用遥感获得的植被光谱特性间接监测土壤理化性质成为可能。NDVI(归一化植被指数)在监测地表植被覆盖中发挥着重要作用。水田植被覆盖种类单一,其它影响因素少,这使得通过NDVI监测水稻长势,间接监测稻田土壤肥力状况变得可行。利用遥感技术监测土壤肥力状况,正好弥补了传统土壤监测的缺陷。遥感技术以其实时、持久、数据量大、观测范围广等优点在土壤监测中发挥了重要作用,成为土壤各种元素含量监测和时空动态分析的重要技术手段,突破了传统的地面点状监测方法。遥感技术能清楚地反映出土壤一些理化性质的空间分布特征,利用多时相的遥感数据可对同一地区的土壤进行动态监测,防止施肥过度导致的面源污染或施肥不足产生的地力衰竭,为合理的配方施肥提供决策依据。本研究以南京溧水县农区水稻田作为实验区,以研究区水稻的主要遥感影象指数——归一化植被指数(NDVI)作为研究对象。采用相关性回归分析方法将地面定位微观监测数据和遥感技术宏观监测数据有机地对接起来,旨在探讨研究区水田土壤的主要营养元素——全氮在中巴地球资源2号卫星(CBERS-2,简称中巴-2)数据中的敏感波段或波段组合,研究建立农区水田土壤全氮含量的遥感监测估算模型。最后将所建的模型应用于水田肥力的估测和研究区水田土壤全氮含量二级图的划分。本文的创新点在于利用中巴-2号卫星的CCD的归一化植被指数对南京溧水县水田的土壤全氮进行估测,在此基础上做了研究区的氮素等级分布图,并对关于遥感估测土壤元素技术体系的建立进行了讨论和初探。通过全文研究,得出结论如下：1)水稻分蘖期中巴-2卫星影象取得的归一化植被指数可以很好的反映水稻长势,间接表现稻田土壤的全氮水平；2)中巴-2卫星影象的CCD数据可以实现对稻田土壤全氮的半定量估测；3)在归一化植被指数影象的基础上可以实现对研究区稻田土壤全氮水平的区划,并可成图；4)遥感估测土壤元素技术体系的建立基本是可行的。基于上述研究,可通过进一步的研究提高估测模型的精度,进而达到精确监测。通过对土壤其它元素含量和物理性质的遥感监测研究,建立并完善遥感估测土壤元素的技术体系。更多还原

【Abstract】 The remote sensing is important in getting the fertility level of the cropland and plays an important role in the the fertilization. The Normalized Difference Vegetation Index (NDVI) has made great contribution in the monitoring of the earth vegetation cover. The kind of the vegetation cover in the paddy field is single, with few other affecting factors which makes monitoring the soil fertility indirectly with the monitoring of the rice growth vigour using NDVI. The paper focused on the monitoring of the soil fertility of the paddy soil in the county of Lishui, Nanjing city using the NDVI of the CBERS-2 CCD image. And the research showed that is possible. And the grade distribution figure of the total Nitrogen content of the study area is made, which makes the monitoring of the total Nitrogen in the paddy field possible with remote sensing.Soil is the material foundation and the headspring of human survival for the people living on the earth. As the agency which is very important for the crop growth and provides food for the world, soil is the plant base for its growth reproduction and biological production. And the content of various elements in the soil is the main factor affecting the soil fertility and deciding whether the srop has a high yield or not Therefor, understand the soil fertility and realize the crop restriction factor according to the basic nutritional requirement of the crop production is critical to the rational fertilization and the income increase of the crop.As there are spatial distribution differences of the soil fertility in the cropland, the traditional indistinctive fertilization may lead to agricultural no point pollution because of over fertilization in some area whereas can cause soil fertility exhaustion in other places. Besides, the content of various elements in the soil is a relative dynamic state which is a changing process. So it is necessary to real time monitor all kinds of nutrient elements in the soil because the soil fertility in the cropland must to be mastered if formulated fertilization want to be executed. The traditional monitoring to the soil fertility usually implemented through the steps of field sampling, air drying, soil grinding and the chemical analysis in the laboratory. On one hand, it will wast a mass of labours and material resources; on the other hand,the period of monitoring the soil fertility will last for long and that will make a bad time effectiveness. Therefore the peasants can not master the information of the soil fertility in the cropland effectively in time and it is hard to take efficient measures to execute rational fertilization for removing the soil restriction factor to the crops. So it is urgent to find a method to monitor the soil fertility quickly, effectively with low input.The remote sensing technology is a comprehensive detective technology which springed up in the sixties of the 20th century and then developed rapidly. It is a process to conclude earth surface parameters by the measure of the electromagnetic radiation obtained from the ground surface. As the common ground object, soil is very sensitive to the electromagnetic radiation of specific waveband. There is a remarkable correlation between soil electromagnetic radiation in some waveband and some physical-chemical parameters of the soil which is the internal mechanism of monitoring soil quality with remote sensing technology. But the soil in the cropland was usually covered by vegetation and the duration of the bare soil was short that affected the application of remote sensing technology in the soil fertility monitoring in the croplang. However, scholars at home and abroad have make remarkable achievements on the monitoring of crop growth status and its nutritional conditions. Because different physico-chemical properties of the soil and different levels of fertilization will affect the crop growth which reflects on the different electromagnetic radiation of the crop canopy finally, the existence of this mechanism make it possible to indirectly monitor the soil physico-chemical properties through vegetation spectrum properties obtained by remote sensing. The Normalized Difference Vegetation Index (NDVI) has made great contribution in the monitoring of vegetation coverage on the land surface. Because the single kind vegetation coverage in the paddy land is scarcely affected by other factors, it is feasible to indirectly monitor the soil fertility in the paddy land through monitoring the rice growth by NDVI.The application of remote sensing in the monitoring of the soil fertility status just can make up the defects of the traditional soil monitoring. Remote sensing technology has played an important role in the soil monitoring for its advantages such as the real time, persistence, large volume of data, wide range of observation and so on. The spatial distribution of the soil physical-chemical properties can be reflected using the remote sensing, so the soil in the same area could be monitored by using multi-temporal remote sensing images that will prevent the no_point pollution caused by over fertilization in some area and the soil fertility exhaustion in other places which can provide evidence for the rational fertilization.The research took paddyland of the rural area in Lishui County, Nanjing as research region, took the main index of the remote sensing image—NDVI as the research object. The correlation and regression analysis was make to connect organically the relationship between the micro monitoring data on the ground and the macroscopically monitoring data of the remote sensing on the purpose of probing into the sensitive wave band or wave band association of the main nutrient element of the soil-total nitrogen in rural area from China and Brazil Earth Resource Satellite 2 data (CBERS-2 data), studied the establishment of remote sensing monitoring speculating model of soil quality in rural district. At last, the established model was applied to monitor the soil fertility of the paddyland and the map of the total nitrogen content of the paddy soil in the research region was also made.The innovative point of this paper is that the soil total nitrogen content of the paddyland in Lishui County, Nanjing was estimated by the Normalized Difference Vegetation Index (NDVI) of the CCD data from the CBERS-2; and on this basis the grade distribution of the soil total nitrogen content in the research area was made; further more, the preliminary study of the founding of the technology system of the estimation of the soil elements in the rice field using remote sensing was also discussed. This paper reached the following conclusions by the study:(1)The NDVI of the CCD data from the CBERS-2 in the tillering stage of the rice can reflect the rice growth propertily and the total nitrogen level of the soil in the paddyland can also be indirectly reflected through that; (2)the semi-quantitative estimation of the soil total nitrogen in the paddyland could be realized using the CCD data from the CBERS-2; (3) regionalization mapping of the soil total nitrogen in the paddyland of the research region also can be achieved on the basis of the NDVI image; (4) it is basicly feasible to found the technology system of the estimation of the soil elements with remote sensing technology. On the basis of the research above, further research can be done to improve the precision of the estimation model,then the accurate monitoring to the soil elements can be realized. Furthermore found and perfect the technology system of the estimation of the soil elements with remote sensing technology through monitoring the content of other elements in the soil.更多还原