Abstract

Abstract. We use 2011–2019 aerosol optical depth (AOD) observations from the Geostationary Ocean Color Imager (GOCI) instrument over East Asia to infer 24 h daily surface fine particulate matter (PM2.5) concentrations at a continuous 6 × 6 km2 resolution over eastern China, South Korea, and Japan. This is done with a random forest (RF) algorithm applied to the gap-filled GOCI AODs and other data, including information encoded in GOCI AOD retrieval failure and trained with PM2.5 observations from the three national networks. The predicted 24 h GOCI PM2.5 concentrations for sites entirely withheld from training in a 10-fold cross-validation procedure correlate highly with network observations (R2 = 0.89) with a single-value precision of 26 %–32 %, depending on the country. Prediction of the annual mean values has R2 = 0.96 and a single-value precision of 12 %. GOCI PM2.5 is only moderately successful for diagnosing local exceedances of the National Ambient Air Quality Standard (NAAQS) because these exceedances are typically within the single-value precisions of the RF and also because of RF smoothing of extreme PM2.5 concentrations. The area-weighted and population-weighted trends of GOCI PM2.5 concentrations for eastern China, South Korea, and Japan show steady 2015–2019 declines consistent with surface networks, but the surface networks in eastern China and South Korea underestimate population exposure. Further examination of GOCI PM2.5 fields for South Korea identifies hot spots where surface network sites were initially lacking and shows 2015–2019 PM2.5 decreases across the country, except for flat concentrations in the Seoul metropolitan area. Inspection of the monthly PM2.5 time series in Beijing, Seoul, and Tokyo shows that the RF algorithm successfully captures observed seasonal variations in PM2.5, even though AOD and PM2.5 often have opposite seasonalities. The application of the RF algorithm to urban pollution episodes in Seoul and Beijing demonstrates high skill in reproducing the observed day-to-day variations in air quality and spatial patterns on the 6 km scale. A comparison to a Community Multiscale Air Quality (CMAQ) simulation for the Korean peninsula demonstrates the value of the continuous GOCI PM2.5 fields for testing air quality models, including over North Korea, where they offer a unique resource.

Highlights

  • Exposure to outdoor fine particulate matter (PM2.5; less than 2.5 μm in diameter) is a global public health issue, accounting for 8.9 million deaths in 2015 (Burnett et al, 2018)

  • We show results for the class 2 National Ambient Air Quality Standards (NAAQS) in eastern China and for both pre-2018 and post-2018 NAAQS for South Korea because all observed grid cells exceed the new annual NAAQS of 15 μg m−3. c Percentage of site days (24 h standard) or site years exceeding the NAAQS. d Percent of detection (POD) defined as the percentage of exceedances successfully detected. e False alarm ratio (FAR) defined as the percentage of predicted exceedances that did not occur. f Equitable threat score (ETS), which is defined as the ability of the random forest (RF) to predict exceedances beyond random chance

  • We focus on South Korea here because it demonstrates how Geostationary Ocean Color Imager (GOCI) PM2.5 adds considerable information to a region that already has relatively good network coverage, including the detection of PM2.5 hot spots missing from the network, such as the Iksan region on the west coast of South Korea in 2015 that was subsequently added to the network by 2019

Read more

Summary

Introduction

Exposure to outdoor fine particulate matter (PM2.5; less than 2.5 μm in diameter) is a global public health issue, accounting for 8.9 million deaths in 2015 (Burnett et al, 2018). We apply an RF algorithm to 2011–2019 GOCI AOD data to construct a continuous dataset of 24 h PM2.5 concentrations at a 6 × 6 km resolution for eastern China, South Korea, and Japan trained with surface network data. This is a larger spatial domain than has been attempted in previous studies. We ensure continuity by using gap-filled AOD, calculated by blending a CTM simulation with statistical interpolation, along with a parameter characterizing the length scale of the interpolation as inputs to the RF algorithm This strategy maximizes the training set size and allows the RF to determine a strategy to handle information encoded by retrieval failure. We exploit the continuity of the dataset to determine trends of PM2.5 air quality in East Asia over the past half decade

GOCI AODs
Meteorological and geographical predictor variables
AOD gap-filling
Random forest algorithm
Accuracy and precision of RF predictions
Urban-scale pollution events
Conclusions
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call