基于监测指标和同化策略优化的水稻重金属胁迫评估

发布时间：2019-06-26 17:44

【摘要】：重金属污染严重影响农田生态系统健康,由此产生的食品安全问题已威胁到人类健康。农田重金属污染具有滞后性、隐蔽性以及不可逆性等特点,采用传统方法难以实时监测与评估污染情况。因此,利用遥感手段实现农田重金属污染有效监测对食品安全及环境保护具有重要意义。目前遥感监测方法在数据、方法、监测指标等方面仍需进一步优化与提高,以解决时空不连续、监测精度差等问题。本文以湖南省株洲市水稻田为研究区,通过获取和采集相关遥感影像数据及作物生理生化参数、气象参数等实测数据,从监测方法和监测指标两方面改进和优化以遥感和作物生长模型为主要依托的水稻重金属污染监测框架,解决现有研究中时间不连续、监测精度低、数据不理想等问题。本文依据敏感性分析及相关研究结果,确定水稻根重(Weight of root,WRT)为重金属胁迫敏感参数,进而构建以水稻地上生物量(Biomass,BM)为同化目标变量的雷达数据与改进WOFOST(World Food Study)模型同化框架,通过优化嵌入的胁迫系数fCVF和fDTGA,模拟水稻WRT并计算监测指标RMR(Root Mass Ratio)。结果表明,RMR能够削减不同实验区因背景环境因素不同造成的水稻重金属胁迫响应被高估的情况,其在时间和空间尺度上与胁迫系数具有较高一致性。同时,SAR在同化框架中的成功应用对于多云雨地区的作物长势监测具有重要意义。为提高同化模拟精度,本文基于多源遥感数据(Sentinel-1雷达数据及GF-1多光谱数据)与改进后WOFOST模型同化框架,从机理角度探索并构建了基于不同同化目标的同化策略。通过对比不同策略方案下作物WRT模拟精度,评价并选择最优方案进行作物生理生长参数模拟,进而计算监测指标以实现重金属胁迫等级监测与评估。结果表明,基于BM和WRT的双目标参量同化策略具有更高的同化精度,不同胁迫程度下两研究区实测值与模拟值RPE均低于15%,而单目标同化策略RPE高于20%。研究表明,基于双目标同化策略和RMR的水稻重金属胁迫监测框架能够削弱背景环境因素影响,有效提高模拟和监测精度,改善同化过程中遥感数据数量和质量不佳的问题。
[Abstract]:Heavy metal pollution seriously affects the health of farmland ecosystem, and the resulting food safety problems have threatened human health. Heavy metal pollution in farmland has the characteristics of lag, concealment and irreversibility, so it is difficult to monitor and evaluate the pollution in real time by traditional methods. Therefore, it is of great significance to realize the effective monitoring of heavy metal pollution in farmland by remote sensing for food safety and environmental protection. At present, remote sensing monitoring methods still need to be further optimized and improved in the aspects of data, methods, monitoring indexes and so on, in order to solve the problems of time-space discontinuity and poor monitoring accuracy. Taking rice field in Zhuzhou City, Hunan Province as the study area, by obtaining and collecting the relevant remote sensing image data, crop physiological and biochemical parameters, meteorological parameters and other measured data, this paper improves and optimizes the monitoring framework of rice heavy metal pollution based on remote sensing and crop growth model from two aspects of monitoring method and monitoring index, and solves the problems of time discontinuity, low monitoring accuracy and unsatisfactory data in the existing research. Based on sensitivity analysis and related research results, rice root weight (Weight of root,WRT was determined as a sensitive parameter of heavy metal stress, and then radar data and improved WOFOST (World Food Study) model assimilation framework with rice aboveground biomass (Biomass,BM) as assimilation target variable were constructed. Rice WRT was simulated by optimizing embedded stress coefficients fCVF and fDTGA, and monitoring index RMR (Root Mass Ratio). Was calculated. The results showed that RMR could reduce the overestimation of heavy metal stress responses caused by different background and environmental factors in different experimental areas, and it was highly consistent with the stress coefficient on the temporal and spatial scales. At the same time, the successful application of SAR in assimilation framework is of great significance for crop growth monitoring in cloudy and rainy areas. In order to improve the accuracy of assimilation simulation, based on multi-source remote sensing data (Sentinel-1 radar data and GF-1 multispectral data) and improved WOFOST model assimilation framework, the assimilation strategy based on different assimilation targets is explored and constructed from the point of view of mechanism. By comparing the accuracy of crop WRT simulation under different strategy schemes, the optimal scheme was evaluated and selected to simulate the physiological growth parameters of crops, and then the monitoring indexes were calculated to realize the monitoring and evaluation of heavy metal stress levels. The results show that the two-objective parameter assimilation strategy based on BM and WRT has higher assimilation accuracy. Under different stress degrees, the measured and simulated values of RPE are lower than 15%, while the RPE of single-objective assimilation strategy is higher than 20%. The results show that the monitoring framework of heavy metal stress in rice based on double-objective assimilation strategy and RMR can weaken the influence of background environmental factors, effectively improve the accuracy of simulation and monitoring, and improve the poor quantity and quality of remote sensing data in the process of assimilation.
【学位授予单位】：中国地质大学(北京)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：X87;X56

【相似文献】