水稻重金属胁迫生理功能变化遥感分析方法研究

发布时间：2018-07-05 10:03

本文选题：水稻重金属胁迫 + 遥感分析　；参考：《中国地质大学(北京)》2017年博士论文

【摘要】：运用遥感技术快速、准确监测大范围作物重金属污染胁迫对我国粮食安全和耕地保护具有重要意义。已有研究多是在作物生理要素特征或重金属含量与光谱参数之间建立经验模型,从作物生理功能特征角度进行的研究则少见报道。另外,研究使用的光谱多集中在可见光/近红外波段范围,少有研究从胁迫的热红外特征考虑。本文选取湖南省株洲地区水稻种植区作为研究区,采集水稻关键生长期的重金属含量数据、叶面积指数(LAI)数据、冠层温度和气象数据,并获取环境卫星CCD数据、Landsat 8数据和高分一号数据。从生理功能变化角度对水稻重金属胁迫遥感监测方法进行分析,注重机理性和理论模型研究。本文的研究工作与主要结论如下:(1)采用具有较强机理性的WOrld FOod STudies(WOFOST)作物生长模型与遥感数据同化的方法获取重金属胁迫下水稻光合作用和干物质转化效率两种生理功能的变化情况,作为评估重金属胁迫水平的指标。在生长模型模拟日总CO2同化速率及干物质转化系数的位置分别嵌入两个胁迫因子,利用HJ-1 CCD数据反演得到的水稻LAI与WOFOST模型输出的LAI作为耦合点,基于粒子群优化算法求取胁迫因子值。通过分析两个胁迫因子对根、茎、叶、穗等器官生物量累积的影响权重间接推算其对胁迫的影响权重,并构建综合胁迫指数。基于少量样本点水稻两个生长时段的综合胁迫指数构建针对不同胁迫水平的判别规则,并通过逐像元运行同化方法实现大范围水稻重金属胁迫水平判别,准确率达90%以上。(2)分析重金属胁迫下水稻植株生理功能改变引起的冠层温度变化,探讨冠层温度的影响机制及水稻生长和胁迫的极限状态,基于物理能量平衡公式构建“冠气温差-LAI理论三角模型”,并提出归一化重金属胁迫指数(HMSI)。首先在田间尺度对模型进行适用性验证,然后协同热红外遥感与光学遥感数据实现株洲县域内模型的区域验证与示范应用。(3)将表征生理功能变化的综合胁迫指数与根部干重作为遥感监测重金属胁迫的目标特征集,并基于分形理论探讨监测胁迫水平合适的空间分辨率。目标特征的异质性越大越能完整刻画特征变量的微小变化,对应的分维数也越大。因此,对应最高分维数的空间分辨率为最佳分辨率。结果显示,4m至32m为合适的空间分辨率,8m是最佳选择。
[Abstract]:Rapid and accurate monitoring of heavy metal pollution stress by remote sensing is of great significance to food security and cultivated land protection in China. Most of the previous studies have established empirical models between the physiological characteristics of crops or heavy metal contents and spectral parameters, but few studies have been carried out from the perspective of physiological functional characteristics of crops. In addition, most of the spectra used in the study focused on the visible / near infrared band, and few studies were conducted on the thermal infrared characteristics of stress. In this paper, the data of heavy metal content, leaf area index (Lai), canopy temperature and meteorological data were collected from the rice growing area in Zhuzhou region of Hunan province, and the data of heavy metal content, leaf area index (Lai), canopy temperature and meteorological data were collected in the key growing period of rice. The Landsat 8 data and the high score 1 data are obtained from the CCD data of the environmental satellite. The remote sensing monitoring method of heavy metal stress in rice was analyzed from the point of view of physiological function change. The research work and main conclusions are as follows: (1) using WOrld FOod STudies (WOFOST) crop growth model and remote sensing data assimilation method to obtain photosynthesis and dry matter conversion efficiency of rice under heavy metal stress. (1) two physiological methods, WOrld FOod STudies (WOFOST), are used to obtain photosynthesis and dry matter conversion efficiency of rice under heavy metal stress. Changes in function, It is used as an index to evaluate the level of heavy metal stress. Two stress factors were embedded in the position of the total CO _ 2 assimilation rate and the dry matter conversion coefficient of the growth model, respectively, and the Lai output from the rice Lai and WOFOST model obtained by HJ-1 CCD data was used as the coupling point. The stress factor is calculated based on particle swarm optimization algorithm. The influence weight of two stress factors on biomass accumulation of root, stem, leaf, ear and other organs was calculated indirectly and the comprehensive stress index was constructed by analyzing the influence weight of two stress factors on the biomass accumulation of root, stem, leaf, ear and so on. Based on the comprehensive stress index of two growing stages of rice in a small number of samples, the discriminant rules for different stress levels were constructed, and the heavy metal stress levels of large range rice were distinguished by pixel operation assimilation method. The accuracy was over 90%. (2) the changes of canopy temperature caused by physiological function changes of rice plants under heavy metal stress were analyzed, and the influence mechanism of canopy temperature and the limit state of rice growth and stress were discussed. Based on the physical energy balance formula, the "canopy temperature difference-Lai theoretical triangle model" was constructed, and the normalized heavy metal stress index (HMSI) was proposed. Firstly, the applicability of the model is verified at the field scale. Then the regional verification and demonstration application of the model in Zhuzhou county were realized by using thermal infrared remote sensing and optical remote sensing data. (3) the comprehensive stress index and root dry weight, which represented the physiological function changes, were taken as the target feature set of remote sensing to monitor heavy metal stress. Based on fractal theory, the spatial resolution of monitoring stress level is discussed. The greater the heterogeneity of the target feature, the more the fractal dimension of the feature variable can be described completely. Therefore, the spatial resolution corresponding to the highest fractal dimension is the best resolution. The results show that the optimum spatial resolution is from 4 m to 32 m.
【学位授予单位】：中国地质大学(北京)
【学位级别】：博士
【学位授予年份】：2017
【分类号】：X87

【相似文献】