Abstract. In this study, three methods, i.e., the random forest (RF) algorithm, boosted regression trees (BRTs) and the improved complete ensemble empirical-mode decomposition with adaptive noise (ICEEMDAN), were adopted for investigating emission-driven interannual variations in concentrations of air pollutants including PM2.5, PM10, O3, NO2, CO, SO2 and NO2 + O3 monitored in six cities in South China from May 2014 to April 2021. The first two methods were used to calculate the deweathered hourly concentrations, and the third one was used to calculate decomposed hourly residuals. To constrain the uncertainties in the calculated deweathered or decomposed hourly values, a self-developed method was applied to calculate the range of the deweathered percentage changes (DePCs) of air pollutant concentrations on an annual scale (each year covers May to the next April). These four methods were combined together to generate emission-driven trends and percentage changes (PCs) during the 7-year period. Consistent trends between the RF-deweathered and BRT-deweathered concentrations and the ICEEMDAN-decomposed residuals of an air pollutant in a city were obtained in approximately 70 % of a total of 42 cases (for seven pollutants in six cities), but consistent PCs calculated from the three methods, defined as the standard deviation being smaller than 10 % of the corresponding mean absolute value, were obtained in only approximately 30 % of all the cases. The remaining cases with inconsistent trends and/or PCs indicated large uncertainties produced by one or more of the three methods. The calculated PCs from the deweathered concentrations and decomposed residuals were thus combined with the corresponding range of DePCs calculated from the self-developed method to gain the robust range of DePCs where applicable. Based on the robust range of DePCs, we identified significant decreasing trends in PM2.5 concentration from 2014 to 2020 in Guangzhou and Shenzhen, which were mainly caused by the reduced air pollutant emissions and to a much lesser extent by weather perturbations. A decreasing or probably decreasing emission-driven trend was identified in Haikou and Sanya with inconsistent PCs, and a stable or no trend was identified in Zhanjiang with positive PCs. For O3, a significant increasing trend from 2014 to 2020 was identified in Zhanjiang, Shenzhen, Guangzhou and Haikou. An increasing trend in NO2 + O3 was also identified in Zhanjiang and Guangzhou and an increasing or probably increasing trend in Haikou, suggesting the contributions from enhanced formation of O3. The calculated PCs from using different methods implied that the emission changes in O3 precursors and the associated atmospheric chemistry likely played a dominant role than did the perturbations from varying weather conditions. Results from this study also demonstrated the necessity of combining multiple decoupling methods in generating emission-driven trends in atmospheric pollutants.
Text S1. Description of the method calculating the range of the deweathered percentage changeTo calculate the range of the deweathered percentage change (DePC) of an air pollutant in any given two years, five steps were designed. The hourly average PM 2.5 concentrations in Guangzhou from 2014 to 2020 were used as an example below (the code can be downloaded from https://pypi.org/project/DePC/).
Step 1: Construction of a dataset with equal size in each yearIn each year from 2014 to 2020 (with i=1, 2, … and 7 representing the first, second, … and the seventh year), there are N i data points of hourly PM 2.5 mass concentration.Rearrange all data points in each year from the smallest to the largest values to generate an array of data, A i (j), with j=1 to N i .For any two selected years: A i (j) with j=1 to N i and A i ' (j) with j=1 to N i ', set N s = Min (N i , N i '). Then convert A i (j) with j=1 to N i into B i (j) with j=1 to N s , and convert A i ' (j) with j=1 to N i ' into B i' (j) with j=1 to N s .If N i = N i ', B i (j)=A i (j) and B i' (j)=A i' (j) with j=1 to N s . If N i > N i ' (N i '=N s ), the difference (𝑛𝑛) of N i -N s is calculated together with the quotient of N i /n for filtering data.If N i is divisible by 𝑛𝑛, N i /n equals to an integer as 𝑘𝑘. Then B i (j) is calculated as below:1) with j=1 to k-2, B i (j) = A i (j), and with j=k-1, B i (k-1) = (A i (k-1) + A i (k))/2; 2) with j=k to 2*k-3, B i (j) = A i (j+1),and with j=n*k-n. B i (n*k-n) = (A i (n*k-1) + A i (n*k))/2. Note that if N i is not divisible by 𝑛𝑛, round it down to an integer as 𝑘𝑘, i.e., 𝑘𝑘 = [𝑁𝑁 𝑖𝑖 /𝑛𝑛].Then use k 1 =k and k 2 =k+1 to calculate corresponding n 1 and n 2 to meetIf n 1 =n 2 , use k=k 1 in 1) and k=k 2 in 2) and repeat the replacement through the conversion, else choose n'=|n 1 -n 2 |.There are two scenarios, a) when n 1 >n 2 , use k=k 1 in 1), 2), …, n'), n'+1), n'+3), …, n-1) and k=k 2 in n'+2), n'+4), …, n) to process the replacement; b) when n 1
Abstract. In this study, three methods including the random forest (RF) algorithm, boosted regression trees (BRTs) and the improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) were adopted for investigating emission-driven interannual variations in concentrations of air pollutants including PM2.5, PM10, O3, NO2, CO, SO2 and (NO2+O3) monitored in six cities in south China from May 2014 to April 2021. The first two methods were used to calculate the deweathered hourly concentrations, and the third one was used to calculate decomposed hourly residuals. To constrain the uncertainties in the calculated deweathered or decomposed hourly values, a self-developed method was applied to calculate the range of the deweathered percentage changes (DePCs) of air pollutant concentrations in annual scale. Emission-driven trends and emission-driven percentage changes (PCs) during the whole seven-year period were generated with the four methods being applied to analyzing the data. The consistency in the trends between the RF-deweathered and BRTs-deweathered concentrations and the ICEEMDAN-decomposed residuals of an air pollutant in a city reaches approximately 70 % of all the studied cases, but that in the PCs reaches only approximately 30 % of all the cases. The remaining cases with inconsistent trends and/or PCs indicated large uncertainties produced by one or more of the three methods. The calculated PCs from the deweathered concentrations and decomposed residuals were thus combined with the corresponding range of DePCs calculated from the self-developed method to gain the robust range of DePCs where applicable. Building on the robust ranges, the mitigation effects were discussed.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.