Quantitative design of emergency monitoring network for river chemical spills based on discrete entropy theory
Quantitative design of emergency monitoring network for river chemical spills based on discrete entropy theory
复制标题
基于离散熵理论的河流化学品泄漏应急监测网络定量设计
DOI：
10.1016/j.watres.2018.01.057
复制
发表时间：
2018
期刊：
影响因子：
12.8
通讯作者：
Wang Peng
中科院分区：
文献类型：
--
作者：
Shi Bin;Jiang Jiping;Sivakumar Bellie;Zheng Yi;Wang Peng
现场监测策略对于备灾和流域紧急环境管理至关重要。然而，发展这种技术也是一项极具挑战性的工作。尽管迄今为止的努力和进展，没有明确的指导方针或解决方案，可用于定量设计一个监测网络，以应对河流化学品泄漏事件，除了一般规则的基础上，行政区划或任意插值的例行监测部分。为了解决这一问题，本研究提出了一种新的时空网络设计框架。该框架将污染物输运模型与离散熵理论和谱分析相结合。应用水质模型预测溢油后污染物的时空分布，估算相应的信息传递指数（ITIs）和傅立叶近似周期函数，作为确定采样点和采样时间的关键指标。研究结果表明，该框架能够在溢油风险情景分析的基础上，快速制定应急监测预案，并在突发事件发生时，及时制定应急监测预案。将该框架应用于一个基于示踪剂实验的假设泄漏案例和一个真实的硝基苯泄漏案例，验证了该框架的适用性和有效性。新设计的时空监测网络以相对较低的成本获取了主要的污染信息。它对后续预警和处理以及灾后恢复和评估都显示出明显的益处。基于案例研究，分析了信息技术创新的潜在驱动因素以及该方法的局限性和不确定性。与现有的监测网络设计方法，管理的影响，和一般适用性进行了比较。
Field monitoring strategy is critical for disaster preparedness and watershed emergency environmental management. However, development of such is also highly challenging. Despite the efforts and progress thus far, no definitive guidelines or solutions are available worldwide for quantitatively designing a monitoring network in response to river chemical spill incidents, except general rules based on administrative divisions or arbitrary interpolation on routine monitoring sections. To address this gap, a novel framework for spatial-temporal network design was proposed in this study. The framework combines contaminant transport modelling with discrete entropy theory and spectral analysis. The water quality model was applied to forecast the spatio-temporal distribution of contaminant after spills and then corresponding information transfer indexes (ITIs) and Fourier approximation periodic functions were estimated as critical measures for setting sampling locations and times. The results indicate that the framework can produce scientific preparedness plans of emergency monitoring based on scenario analysis of spill risks as well as rapid design as soon as the incident happened but not prepared. The framework was applied to a hypothetical spill case based on tracer experiment and a real nitrobenzene spill incident case to demonstrate its suitability and effectiveness. The newly-designed temporal-spatial monitoring network captured major pollution information at relatively low costs. It showed obvious benefits for follow-up early-warning and treatment as well as for aftermath recovery and assessment. The underlying drivers of ITIs as well as the limitations and uncertainty of the approach were analyzed based on the case studies. Comparison with existing monitoring network design approaches, management implications, and generalized applicability were also discussed.