Recently, a variety of modeling studies have been conducted to examine the air quality over South Korea during the Korea United States Air Quality (KORUS-AQ) campaign period (May 1 to June 10, 2016). This study investigates the impact of different meteorological initializations on atmospheric modeling results. We conduct several simulations during the KORUS-AQ period using the Weather Research and Forecasting (WRF) model with two different initial datasets, which is FNL of NCEP and ERA5 of ECMWF. Comparing the raw initial data, ERA5 showed better accuracy in the temperature, wind speed, and mixing ratio fields than those of NCEP-FNL. On the other hand, the results of WRF simulations with ERA5 showed better accuracy in the simulated temperature and mixing ratio than those with FNL, except for wind speed. Comparing the nudging efficiency of temperature and wind speed fields, the grid nudging effect on the FNL simulation was larger than that on the ERA5 simulation, but the results of mixing ratio field was the opposite. Overall, WRF simulation with ERA5 data showed a better performance for temperature and mixing ratio simulations than that with FNL data. For wind speed simulation, however, WRF simulation with FNL data indicated more accurate results compared to that with ERA5 data.

Abstract
1. 서 론
2. 연구 방법
    2.1. 모델 설정
    2.2. 실험 설계
    2.3. 관측 자료 및 통계 지표
3. 결과
    3.1. 원시 자료간의 비교
    3.2. 초기 입력 자료별 WRF 수치 모의 결과 비교
4. 결 론
REFERENCES

• 문정혁(부산대학교 지구환경시스템학부) | Jeonghyeok Mun (Division of Earth Environmental System, Pusan National University)
• 전원배(부산대학교 지구환경시스템학부) | Wonbae Jeon (Department of Atmospheric Sciences, Pusan National University)
• 이화운(부산대학교 대기환경과학과) | Hwa Woon Lee (Department of Atmospheric Sciences, Pusan National University) Corresponding author