论文部分内容阅读
多尺度数据整合方法以贝叶斯(Bayesian)层次模型为框架,融合了尺度提升的思想并能够有效地结合马尔科夫链蒙特卡洛抽样算法,它在将所有相关信息进行整合的同时考虑到先验信息的贡献,为不同来源、不同尺度间动态和(或)静态数据的整合提供了有效的模拟框架。本文基于土壤饱和与非饱和水力特性参数空间定量化研究的重要意义,综述了多尺度数据整合方法的发展及其在该研究中的应用现状,展望了它在进一步整合不同尺度的静态和(或)动态数据从而进行土壤水力参数空间定量化中的应用前景,并分析了延拓应用中所面临的理论和技术突破点,为结合专业知识、运用多尺度数据整合方法探究土壤水文学中的关键科学问题提供了理论分析和技术支撑的参考依据。
The multi-scale data integration method, based on the Bayesian hierarchical model, incorporates the idea of scale-up and can effectively integrate the Monte Carlo sampling algorithm of Markov chain. It takes all the related information into account while taking into account The contribution of prior information provides an effective simulation framework for the integration of dynamic and / or static data from different sources and scales. Based on the importance of spatial quantitative research on saturated and unsaturated hydraulic characteristics of soil, this paper reviews the development of multi-scale data integration and its application in this field. It is expected that it will further integrate the static and / ) Dynamic data so as to quantify the spatial parameters of soil hydraulic parameters and to analyze the theoretical and technical breakthrough points in extension applications. In order to explore the key points of soil hydrology in combination with professional knowledge and using multi-scale data integration methods Scientific issues provide a theoretical basis for analysis and technical support.