耦合信息熵和Copula函数的复杂性大河水文站网优化研究

Hydrology laid a foundation for the country and society. Hydrometric network laid a foundation for Hydrology. .Optimal hydrometric network is one of the most complicated issues in hydrology and water resources. The common method is using the stationary stochastic model to simulate the transinformation. The insufficiencies of this method are that the hydrologic process is stationary and it needs many parameters to calibrate. Especially when the length of sample series is short, the error of calibration becomes bigger. .The approach of using the Informational Entropy and/or Copula function on the uncertainty and multivariate distribution issues possesses much special superiority. What’s more is that this approach is very suitable to short hydrologic series. Recently, in the field of Entropy and Copula application in hydrology and water resources, there are so many studies in Europe and U.S.A. While in China, the studies are few. As to building a optimal hydrometric network model using Entropy-based Approach or Copula-based Approach, due to the model building difficulty, there is no such paper published for Chinese river. What’s more is that the overseas interrelated theoretic studies are not applicable for the Yellow River. There is no published article using the hybrid Entropy-Copula-based Approach to hydrometric network design. .Hence, it is the urgent need in China to build such models. Based on the former studies from others and the applicant, the main objective of this project is to build 3 kinds optimal hydrometric network models using Entropy-based Approach, Copula-based Approach and hybrid Entropy-Copula-based Approach. And these 3 models should be applicable for the Yellow River. Thus such study of Chinese river can emulate and lead the international advanced level.

水文是重要基础性工作，水文站网是水文的基础。水文站网优化是水文学最为复杂的问题之一。现普遍做法是在平稳假定下建立随机模型表征信息转移。不足之处是实际水文过程非平稳且模型参数较多，特别在样本序列较短时(我国普遍存在)误差较大，具备可突破研究空间。以信息熵和Copula函数研究不确定性和多变量分布问题具有不可替代优越性，又特别适于短样本序列。近年，欧美在这方面取得了可喜成果。国内研究相比少得多。以熵或Copula构建水文站网优化模型，由于建模难度大，尚无中国水文站网同类成果发表。国外理论探讨因局限性，经申请者试用于黄河，结果很差，更不能直接应用。至于耦合熵和Copula建模，国内外均无成果发表。因此，迫切需要研制上述适合我国复杂性大河的水文站网优化新模型。本项目是申请者以往研究的延续和深化，力图在前人和已有工作基础上，最终创建3类适于黄河复杂水情的水文站网优化新模型，赶超并引领国际先进水平。

规划合理的水文站网能够充分反映水文时空变异特征，更好地揭示水文规律。对水文站网进行优化，即在满足资料精度要求基础上，探索最优站网布局，使之能收集准确详尽的水文信息。这对于提高站网效率、节约站网建设成本具有重要意义。本项目基于信息熵理论和 Copula函数，构建了多套水文气象站网优化理论模型与方法，包括基于信息熵的多目标水文站网优化模型，基于Copula熵的水文站网优化模型，基于克里金和信息熵理论相结合的水文站网优化模型，以及考虑径流预测的多目标雨量站网设计方法，并对基于信息熵的四类水文站网优化准则进行应用性评价。利用构建的模型方法评估了黄河流域和长江流域水文气象站站网合理性，对研究区域水文气象站网进行了优化设计，提出了水文气象站网优化设计方案。与其他模型结果相比，本项目提出的一系列模型和方法既可实现对站网信息的定量分析，又能综合多个指标进行评价和优化，具有合理性和有效性。本项目成果可为水文气象站网优化提供科学参考，进一步推动随机水文学、水信息学发展。