Reply on RC2

Jae-Bum Lee

doi:10.5194/gmd-2021-356-ac2

Abstract

Abstract. Despite recent progress of numerical air quality models, accurate prediction of fine particulate matter (PM2.5) is still challenging because of uncertainties in physical and chemical parameterizations, meteorological data, and emission inventory databases. Recent advances in artificial neural networks can be used to overcome limitations in numerical air quality models. In this study, a deep neural network (DNN) model was developed for a 3âd forecasting of 6âh average PM2.5 concentrations: the day of prediction (D+0), 1âd after prediction (D+1), and 2âd after prediction (D+2). The DNN model was evaluated against the currently operational Community Multiscale Air Quality (CMAQ) modeling system in South Korea. Our study demonstrated that the DNN model outperformed the CMAQ modeling results. The DNN model provided better forecasting skills by reducing the root-mean-squared error (RMSE) by 4.1, 2.2, and 3.0âÂµgâmâ3 for the 3 consecutive days, respectively, compared with the CMAQ. Also, the false-alarm rate (FAR) decreased by 16.9â%p (D+0), 7.5â%p (D+1), and 7.6â%p (D+2), indicating that the DNN model substantially mitigated the overprediction of the CMAQ in high PM2.5 concentrations. These results showed that the DNN model outperformed the CMAQ model when it was simultaneously trained by using the observation and forecasting data from the numerical air quality models. Notably, the forecasting data provided more benefits to the DNN modeling results as the forecasting days increased. Our results suggest that our data-driven machine learning approach can be a useful tool for air quality forecasting when it is implemented with air quality models together by reducing model-oriented systematic biases.

Highlights

The forecasts are announced four times daily (at 05:00, 11:00, 17:00, and 23:00 (LST)), and the predicted daily average PM2.5 concentrations are represented via four different air quality index (AQI) categories in South Korea: 40 good (PM2.5 ≤ 15 μgm-3), moderate (16 μgm-3 ≤ PM2.5 ≤ 35 μgm-3), bad (36 μgm-3 ≤ PM2.5 ≤ 75 μgm-3), and very bad (76 μgm-3 ≤ PM2.5)
5 Conclusion This study aimed to develop a deep neural network (DNN) model for predicting the 6-hour average PM2.5 concentration for three days 350 (D+0 to D+2) using the DNN algorithm based on observation, weather forecast, and PM2.5 concentration forecast data
The DNN-OBS performed poorly compared to the Community Multiscale Air Quality (CMAQ) model, with root-mean-squared error (RMSE)-increases of 2.8 μgm-3 and 3.3 μgm-3 and sharp index of agreement (IOA)-decreases of 0.46 and 0.58, for D+1 and D+2, respectively

Summary

Introduction

Fine particulate matter (PM2.5) refers to tiny particles or droplets in the atmosphere that exhibit an aerodynamic diameter of less than 2.5 μm. When the forecasts were based on the CMAQ model, the accuracy (ACC) of the daily forecast for the following day (D+1) in Seoul, South Korea, over the three-year period from 2018 to 2020 was 64%, and the prediction accuracy for the high-concentration categories ("bad" and "very bad") was 69%. It is to be noted that previous studies concerning the prediction of PM2.5 concentrations using ANNs primarily developed and evaluated 60 models for predicting the daily average concentration within a 24-hour period based solely on observation data. The daily and 6hour average prediction performance was comparatively evaluated against that of the CMAQ model currently operational for such predictions. The effect of the training data on the daily prediction performance of the DNN model was quantitatively analyzed 65 via three experiments that used different configurations of the training data in terms of predictive data from numerical models as well as observation data

Objectives

Results

Conclusion