面向土壤呼吸监测的控制域覆盖与多维信息融合计量研究

Soil respiration is one of the important factors affecting the atmospheric CO2 concentration, so objective evaluation on the influence of soil respiration to climate change has become one of the key problems about the global carbon cycle. Because of the temporal and spatial heterogeneity of soil respiration carbon flux, current methods of selecting a few feature points and measuring asynchronously cannot accurately monitoring the dynamic of soil respiration. In this project we introduced wireless sensor network technology into the monitoring of dynamic of regional soil respiration. Using a single point of soil respiration carbon flux measurement device developed by ourselves that support wireless sensor network communication, we are going to do such researches as following: (1) adaptive incremental deployment for wireless sensor networks with control domain coverage, that is, pre-deploying positions are recommended with a greedy method and are evaluated by a substituting method according to the measured data from deployed nodes, and the evaluated result are used to determine whether the change pre-deployment to determined deployment; (2) measurement scheduling based on compressive sensing, that is, representation basis matrices to sparsify object data are selected taking advantage of the spatial-temporal correlation of data so as to recover object data from a few numbers of measurements; (3) error detection and recovery of sensing data with the cooperation of multidimensional data, and regional soil respiration carbon flux calculation with the fusion multidimensional sensing data. The results of this research can be used to realize a large scale, full range of accurate measurement of regional soil respiration carbon flux, and provide accurate and credible regional soil respiration carbon flux data for such applications as carbon emission reduction, carbon trading or precision agriculture.

土壤呼吸是影响大气CO2浓度的重要因素，客观评价土壤呼吸对气候变化的影响已成为全球碳循环研究的核心问题之一。由于土壤呼吸碳通量的时空异质性，传统选取少量特征点依次测量的方法无法准确监测区域土壤呼吸动态。本课题将无线传感网技术引入区域土壤呼吸动态监测，利用自主研制的支持无线传感网通信的单点土壤呼吸碳通量测量设备，研究(1)无线传感网控制域覆盖的环境自适应增量部署：根据已部署节点的实测数据，用贪心法推荐新增采样点的近似最优位置，然后对推荐位置进行实测评价，逐步增加采样点；(2)基于压缩感知的测量调度：利用数据时空相关性，选择便于稀疏化目标数据的表达基，以便用少量测量恢复目标数据；(3)多维协同的感知数据错误检测与修复和多维感知数据融合的区域土壤呼吸碳通量计算。本课题的成果可实现区域土壤呼吸碳通量的大尺度、全方位准确计量，并为节碳减排、碳交易或设施农业等应用提供准确可信的区域土壤呼吸碳通量数据。

土壤呼吸是影响大气 CO2 浓度的重要因素，客观评价土壤呼吸对气候变化的影响已成为全球碳循环研究的核心问题之一。由于土壤呼吸碳通量的时空异质性，传统选取少量特征点依次测量的方法无法准确监测区域土壤呼吸动态。本课题将无线传感网技术引入区域土壤呼吸动态监测，利用自主研制的支持无线传感网通信的单点土壤呼吸碳通量测量设备，围绕面向土壤呼吸监测的控制域覆盖与多维信息融合计量的研究主题，对相关基础理论进行了深入研究。在控制域覆盖的环境自适应增量部署策略方面，提出了基于多向插值离散度的空间采样策略、基于湿度分布特征的小尺度土壤碳通量空间采样策略、基于轻量参数时域特征相似性的土壤呼吸监测传感网节点部署策略。在节点的采样调度模型方面，提出了基于压缩感知的分段动态采样调度策略。在基于传感网数据的森林生态系统碳通量计量方面，本课题提出了面向下垫面不均一的森林碳通量监测方法、基于通量源区模型的森林生态系统碳通量信息提取方法等理论方法；这些研究成果为基于传感网技术的土壤呼吸监测系统的构建提供了理论依据,可实现区域土壤呼吸碳通量的大尺度、全方位准确计量，并为节碳减排、碳交易或设施农业等应用提供准确可信的区域土壤呼吸碳通量数据。