Dynamical Networks for Smog Pattern Analysis

TL;DR

This paper introduces a dynamical network model with spontaneous recovery to analyze smog patterns, revealing insights into smog outbreaks, dissipation, and the impact of pollution source distribution on air quality.

Contribution

It is the first to apply dynamical network modeling to smog pattern analysis, providing a new mathematical framework for understanding smog dynamics and control strategies.

Findings

Smog outbreaks and dissipation can be explained by the model.

Internal pollution sources are more critical than external ones.

Moving pollution sources outward can worsen overall air quality under certain conditions.

Abstract

Smog, as a form of air pollution, poses as a serious problem to the environment, health, and economy of the world[1-4] . Previous studies on smog mostly focused on the components and the effects of smog [5-10]. However, as the smog happens with increased frequency and duration, the smog pattern which is critical for smog forecast and control, is rarely investigated, mainly due to the complexity of the components, the causes, and the spreading processes of smog. Here we report the first analysis on smog pattern applying the model of dynamical networks with spontaneous recovery. We show that many phenomena such as the sudden outbreak and dissipation of smog and the long duration smog can be revealed with the mathematical mechanism under a random walk simulation. We present real-world air quality index data in accord with the predictions of the model. Also we found that compared to external causes such as pollution spreading from nearby, internal causes such as industrial pollution and vehicle emission generated on-site are more critical to the health of the whole network. Most strikingly, we demonstrate that spreading out the pollution sources by moving them to nearby suburban areas, a common practice for smog control, will degrade the overall health of system under certain conditions