Abstract:Abstract. A comprehensive understanding of the effects of meteorology, emissions, and chemistry on severe haze is critical in the mitigation of air pollution. However, such an understanding is greatly hindered by the nonlinearity of atmospheric systems. In this study, we developed the quantitative decoupling analysis (QDA) method to quantify the effects of emissions, meteorology, chemical reactions, and their nonlinear interactions on fine particulate matter (PM2.5) pollution by running built-in scenario simulations in each model step. Different from previous methods, the QDA method achieves a fully decomposed analysis of hourly changes in the PM2.5 concentration during pollution events into seven parts, including the pure meteorological contribution (M), the pure emissions contribution (E), the pure chemistry contribution (C), and the interactions among these processes (i.e., ME, MC, EC, and MCE). Via embedding the QDA method into the Weather Research and Forecasting–Nested Air Quality Prediction Modeling System, we employed this method and combined it with the Integrated Process Rate method to study a typical heavy haze episode in Beijing. We evaluate the model performance against in situ meteorological and air quality observations and describe the QDA analytical factors of this case. Results showed that M varied most significantly at different stages of the episode, from 0.21 µg⋅m−3⋅h−1 during the accumulation stage to −11.82 µg⋅m−3⋅h−1 during the removal stage, indicating that the pure meteorological contribution dominated the hourly fluctuation amplitude of the PM2.5 concentration. M acted as the most important cleaner for PM2.5 in non-polluting periods but stopped being effective at this and instead became a contributor in the accumulation stage such that PM2.5 tended to grow rapidly under the superimposed influence of emissions and chemical processes, which would probably mark the beginning of a heavy pollution event. The contribution of E ranged from 0.63 to 0.88 µg⋅m−3⋅h−1 owing to the diurnal variation of emissions. The pure chemical contribution was shown to increase with the level of haze, becoming the largest (0.37 µg⋅m−3⋅h−1) in the maintenance period, which was 25 % higher than during the pre-contamination period. And C+CE made a significant contribution in the accumulation and maintenance stages, indicating that chemical reactions are more important in the polluted period than in other periods. Nonnegligible nonlinear effects exist among the processes of meteorology, emissions, and chemistry on PM2.5 concentrations (−1.83 to 2.44 µg⋅m−3⋅h−1) – something that has generally been ignored in previous studies and during the development of heavy-pollution control strategies. The nonlinear effects are helpful in eliminating the interference of other processes and obtaining a more purified result of the target process and have important indicative significances. The ratio of CE to C is positively correlated with the chemical speed. For precursors like NH3, the smaller value of CE in the most polluted period indicated that NH3 was more deficient, and thus reducing emissions of it in that period would have had the most efficient controlling effect on the PM2.5. This study highlights that the QDA method can be used to realize an in-depth understanding of the effects of adverse meteorological conditions in haze and to judge whether the precursors are excessive or not. Not only can the QDA method provide researchers and policymakers with valuable information for understanding the key factors behind heavy pollution, but it can also help modelers to identify the sources of uncertainties in numerical models.

Quantifying Haze Effect using Air Pollution Index Data

Predicting Particulate Matter (PM2.5) in Malaysia using Multiple Linear Regression and Artificial Neural Network

Assessment of spatial air quality on the East Coast of Peninsular Malaysia utilizing environmetric techniques

Measuring and Modeling Aerosol: Relationship with Haze Events in Shanghai, China

Application of Functional Time Series Model in Forecasting Monthly Diurnal API Curves: A Comparison between Multi-Step Ahead and Iterative One-Step Ahead Approach

Prediction of Particulate Matter (PM10) during High Particulate Event in Peninsular Malaysia using Novel Hybrid Model

Improving the quantification of fine particulates (PM 2.5 ) concentrations in Malaysia using simplified and computationally efficient models

Modelling Malaysia Air Quality Data using Bayesian Structural Time Series Models

Analysis of the Temporal and Spatial Distribution of Haze and Its Influencing Factors in Shanghai

Predict the effect of meteorological factors on haze using BP neural network

Ambient Atmospheric Aerosol Extinction Coefficient Reconstruction from PM2.5 Mass Concentrations and Application to Haze Identification in China

Twenty Years of Air Pollutant Index Trend Analysis in Kuching, Sarawak, Malaysia (2000-2019)

An Intelligent IoT-Cloud-Based Air Pollution Forecasting Model Using Univariate Time-Series Analysis

A comparison of haze pollution variability in China using haze indices based on observations.

Performance Comparison of Malaysian Air Pollution Index Prediction Using Nonlinear Autoregressive Exogenous Artificial Neural Network and Support Vector Machine

Characterization of Haze Episodes and Factors Contributing to Their Formation Using A Panel Model

Predictive Modelling of Air Quality Index (AQI) Across Diverse Cities and States of India using Machine Learning: Investigating the Influence of Punjab's Stubble Burning on AQI Variability

Statistical Model of the Relationship Between Atmospheric Visibility and PM_(2.5) in Shenzhen

Forecasting Daytime Ground-Level Ozone Concentration in Urbanized Areas of Malaysia Using Predictive Models

Association between P M 10 and respiratory diseases admission in peninsula Malaysia during haze

A Quantitative Decoupling Analysis (QDA V1.0) Method for the Assessment of Meteorological, Emission and Chemical Contributions to Fine Particulate Pollution