Utilizing the GOCI Satellite to Estimate Hourly PM2.5 Concentrations in South Korea from 2015 – 2018 Público

Le, Sophia (Spring 2020)

Permanent URL: https://etd.library.emory.edu/concern/etds/1j92g844g?locale=es
Published

Abstract

Air quality in South Korea has been deteriorating in recent years, due to both domestic sources and long-range transport from China. Ground PM2.5 measurements are primarily located in urban environments, resulting in limited spatial and temporal coverage. To address this limitation, AOD retrieved from GOCI was utilized in multiple random forest machine learning models to estimate PM2.5 concentrations. We developed 8 separate random forest machine learning models with time-varying meteorology and spatially fixed land information parameters to be included in the model. A 6 km modeling grid with 6,307 pixels was used to match the parameters to the GOCI retrievals. The average GOCI AOD coverage in the study domain was 48%. The 10-fold cross validation R2 ranged from 0.47 – 0.54 with RMSEs’ ranged between 9.74 – 13.34 µg/m3. The regression slope between observed and predicted hourly concentrations ranged between 1.2 – 1.3. Prediction maps of hourly PM2.5 levels indicate a higher concentration on the western coast of South Korea compared to the eastern coast. We further analyzed an episode of long-range transport from China to South Korea during March 10th and 11th, 2015. Results from those findings revealed evidence of long-range transport given that the western coast has higher concentrations. Overall, our hourly prediction models allow us to understand the spatiotemporal variations of PM2.5 concentrations in South Korea that could not have been done by ground stations alone.

Table of Contents

1.    Introduction.................................. 1

1.1 PM2.5 Background................... 1

1.2 South Korea’s Air Quality and Meteorology..................... 1

1.3 Ground Measurements of PM2.5 in South Korea........ 3

1.4 Air Pollution Studies in Korea Utilizing GOCI AOD....... 3

1.5 Study Objectives..................... 4

2.    Data and Methods......................... 6

2.1 Study Domain......................... 6

2.2 Dataset Description................ 6

2.2.1 GOCI Satellite YAER Version 2 Data............ 6

2.2.2 Ground PM2.5 Measurements............. 7

2.2.3. Meteorology Data......... 7

2.2.4. Ancillary Data.............. 8

2.2.5. Data Integration............ 9

2.3 Model Development............... 9

2.3.1 Random Forest Machine Learning Model........ 10

2.3.2 k-fold Cross Validation 11

3.    Results........................................ 13

3.1 Descriptive Statistics............ 13

3.2 Spatiotemporal AOD Coverage........................................ 14

3.3 Model Performance and Prediction........................ 16

3.4 Long Range Transport.......... 18

4.    Discussion and Conclusion........ 19

5.    References.................................. 22

6.    Tables and Figures...................... 27

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
School
Department
Degree
Submission
Language
  • English
Research Field
Palabra Clave
Committee Chair / Thesis Advisor
Última modificación

Primary PDF

Supplemental Files