基于回归克里格的生态水文无线传感器网络布局优化

在黑河上游八宝河流域建立自动化、时空协同、智能观测的生态水文无线传感器网络,实现分布式的地面观测,对于定量刻画流域尺度时空异质性较强的生态水文要素的动态特征及其不确定性具有重要意义。在观测网设计过程中,节点的空间布局将直接影响到无线传感器网络的观测水平。为准确捕捉流域内关键生态水文要素的时空变异性和场分布,探讨了一种基于回归克里格模型的空间采样布局优化方法,并以地表温度观测网优化为例,应用到八宝河流域生态水文无线传感器网络布局方案设计中。研究结果表明,该优化方法同时考虑了目标变量与环境变量之间的相关关系以及残差在空间上的自相关特征,可以同时优化目标变量的地理空间和属性空间。优化后的无线传感器网络可以较好地捕捉流域内生态水文要素的时空动态特征。

Regression Kriging Model-based Sampling Optimization Design for the Eco-hydrology Wireless Sensor Network

The Babao River Basin is the upstream region of the Heihe River which is the second largest inland river basin in the arid regions of northwest China. To monitor the characteristics of space-time of the eco-hydrological processes in the Babao River Basin, this paper discussed a regression kriging model based sampling optimization method. Land surface temperature, as one of eco-hydrological variables in the Babao River Basin, has been exemplified. The experiment results demonstrate that this sampling optimization method can consider the relationship of target variable and environmental variables and the spatial autocorrelation of regression residuals to obtain the optimization design in the geographic space and attribute space simultaneously. The optimized WSN is more efficient to capture the temporal and spatial variations of the eco-hydrological variables for monitoring the eco-hydrology process in the Babao River Basin.