Dynamic early warning of regional atmospheric environmental carrying capacity

Highlights

• We established an early warning model of atmospheric environmental carrying capacity.

• This early warning model uses a cloud model and Markov chain.

• It can forecast the change trends for atmospheric environmental carrying capacity.

• It can identify and diagnose the potential regional atmospheric environmental risks.

• The results provide evidence to support the formulation of environmental policies.

Abstract

Economic development cannot exceed the maximum amount that the environment can support. Therefore, atmospheric environmental policy should be formulated based on the scientific assessment of regional atmospheric environmental carrying capacity. The establishment of an early warning model of atmospheric environmental carrying capacity can dynamically analyse regional atmospheric environmental carrying capacity, which contributes to discerning the change trend of the regional atmospheric environmental carrying capacity and the risk issue of the regional atmospheric environment. Additionally, it can provide theoretical reference for the formulation of relevant binding and restrictive policies. In this study, according to the daily monitoring data of atmospheric pollutants, we established a dynamic early warning model of regional atmospheric environmental carrying capacity based on the cloud model and Markov chain. The research results show that this model has an excellent early warning capability. Moreover, many regions in China have exceeded the atmospheric environmental carrying capacity, especially in North China and Central China. By 2020, North China and Central China for prediction of region with non-overloading are only 9.09% and 12.50%, respectively. China's regional atmospheric environmental carrying capacity is gradually improving. It is predicted that by 2024, regions with non-overloading in North China and Central China will reach 40.91% and 37.50%, respectively. From the overall aspect, there is currently no risk of serious overload in any region.

Introduction

Currently, the ability of human beings to exploit and utilize nature is increasing, which makes the contradiction between the population, resources and environment serious (Su and An, 2018). Although natural resources in China are plentiful and diverse, the very high population in China means that the amount of resources per capita is low. Meanwhile, the unreasonable and unscientific development and utilization of resources have led to significant resource losses and waste. In the course of consuming natural resources, humans have also seriously polluted the water, soil and air and destroyed the natural environment. Therefore, in many regions of China, the pollution of agricultural water and soil is increasing. Soil erosion and desertification are still intensifying in some regions, and sand-dust storms and fog-haze frequently occur. Especially in developing countries, such as China, with rapid economic development, i.e., the continuous advancement of urbanization and industrialization, a series of ecological and environmental problems, such as water and air pollution, acid rain, haze, energy shortage, and a sharp decline in biodiversity, have also become increasingly prominent (Qiang et al., 2011). These issues have seriously affected China's progress in achieving the sustainable development goals of the United Nations. Additionally, the extensive economic development model, with high energy consumption, high pollution and low efficiency, in China is an important factor that leads to environmental pollution (Wang, 2017; Su and Yu, 2019). In recent years, atmospheric pollution caused by fine particulate matter (PM2.5) has attracted widespread attention. The deterioration of the atmospheric environment caused by the high growth of traditional energy consumption restricts the development of China's social economy (Li et al., 2019). Furthermore, air pollution has been recognized as an important public health issue by the International Agency for Research on Cancer (IARC) of the World Health Organization (WHO) (Niu et al., 2019). The atmospheric environment is a complex system that is influenced by the population, economy, society, etc. (Wang et al., 2018). Hence, air pollution will not only affect economic development but also cause serious harm to human health.

As an index reflecting the interactive relationship between humans and the environment, environmental carrying is usually an important basis for judging the coordinated development of the economy and environment. Environmental carrying capacity refers to the maximum human activity level that the regional environmental system can withstand under specific temporal and spatial conditions (Cheng et al., 2019). In research on environmental carrying capacity, atmospheric environmental carrying capacity is an important dimension. Currently, there are two viewpoints on the definition of atmospheric environmental carrying capacity. One is the volume viewpoint, that is, under the condition of not exceeding the ecosystem limit, the maximum amount of pollutants that the atmospheric environmental ecosystem can bear, the economic scale that can be supported and the population number (Lu et al., 2017). Another is the view of threshold, that is, the maximum load supporting capacity of the atmospheric environmental ecosystem to atmospheric pollutants produced by human activities under certain conditions (Swiader, 2018; Li et al., 2019). Hence, atmospheric environmental carrying capacity, as an effective tool for coordinating the relationship between regional economic development and atmospheric environmental protection, can help with the formulation of environmental protection policies and the planning of the regional economic development structure (Han et al., 2014).

China is actively responding to the environment and climate change. Laws and regulations have been introduced successively, such as the new National Standard for Environmental Air Quality in China (GB 3095-2012), the Action Plan for the Prevention and Control of Air Pollution, and the new Environmental Protection Law. Additionally, China has proposed the concept of ecological civilization, and a well-off society with a good ecological environment should be built by 2020 (Zhang et al., 2018). Furthermore, the “Decision of the Central Committee of the Communist Party of China on Several Major Issues Concerning Comprehensively Deepening Reform” adopted by the Third Plenary Session of the 18th Central Committee of the Communist Party of China clearly stated that a mechanism of monitoring and early warning for the environmental carrying capacity should be established, and restrictive measures should be taken in regions with overload. The purpose of this goal is to ensure that each indicator is within acceptable limits (Zhou et al., 2019a). This definition indicates that China has realized that atmospheric environmental carrying capacity plays an important role in the requirements of regional sustainable development and environmental protection (Guo et al., 2018). Establishing a mechanism of monitoring and early warning for environmental carrying capacity is a major task that requires comprehensively deepening reform (Fan et al., 2015).

The acceleration of the networking process and the continuous advancement of computer science and technology have promoted the rapid development of cloud computing, big data, the internet of things and other fields. The widespread use of mobile sensor devices has led to an explosive growth in perceptual data reflecting the physical world and human social activities (Miao et al., 2018). The massive data generated by the extensive use of sensor devices can help improve the capacity of human understanding of the world and quickly respond to emergencies (Wang et al., 2019). Sensing data have the characteristics of real-time arrival, uninterrupted flow, and fast change, and typically, data with these characteristics are called streaming data (Sahmoud and Topcuoglu, 2020). Streaming data are collected from the internet and include remote sensing data of geographic information systems, sensor data, and environmental monitoring data. The wide application of streaming data provides a new idea for the dynamic management of the environment.

Previous studies have mainly focused on the calculation and evaluation of atmospheric environmental carrying capacity using methods such as the A-P value method, the box model method, and system dynamics modelling (Zhou and Zhou, 2017; Ahmadi et al., 2018). However, few studies have focused on regional atmospheric environmental carrying capacity monitoring or early warning systems. Although the research of Lu et al. (2017) involves an early warning model for atmospheric environmental carrying capacity, their research uses existing air pollutant monitoring data to identify early warnings and cannot be used for prediction analysis. In addition, most of the previous studies are based on a specific scope of research and include too many assumptions and qualifications, which can reduce the applicability and effectiveness of predictions. Importantly, these previous methods can only make short-term forecasts, not long-term trend forecasts.

Hence, this study used the streaming data of environmental monitoring and considered the characteristics of regional atmospheric environmental carrying capacity, including dynamicity, openness, and limit property (Shi et al., 2017). Then, the regional atmospheric environmental carrying capacity model was established based on the cloud model and Markov chain, contributing to the timely evaluation of the regional atmospheric environmental carrying capacity and early warning analysis of possible future environmental risks. Traditional time series prediction methods, such as data analysis, machine learning and neural network methods, are short-term accurate prediction methods. They do not usually consider the issues of time granularity; thus, the results cannot satisfy the actual application requirements. However, the monitoring model of this study can effectively predict the long-term span, and further, the development trend of the regional atmospheric environmental carrying capacity is obtained, which can provide a theoretical basis for the formulation of atmospheric environmental protection policies.

The aim of this paper is to establish a dynamic early warning model of atmospheric environmental carrying capacity, which can be used to quickly judge the regional atmospheric environmental overload situation and diagnose and predict the regional atmospheric environmental development status. In order to improve the prediction accuracy of this model, a method based on cloud model and Markov chain and using stream data is proposed. Then, the feasibility and accuracy of this model in actual test are verified. The research hypothesis of this paper mainly includes the following aspects: 1) The issue of concept drift in this study is ignored, that is, the model can predict with an invariant state transition probability matrix; 2) The results only have five cases: good, attention, lightly overloaded, moderately overloaded, and heavily overloaded; 3) The future sudden risks which can affect atmospheric quality, such as natural disaster, climate change, and so on, aren't considered in this model; 4) The result deviation is caused by mandatory environmental policies that may be issued in the future are also neglected in this model. The results can provide the basis for the development of atmospheric environmental control measures and policies tailored to local conditions.