Sensitivity analysis of the WRF model according to the impact of nudging (e.g., nudging techniques and application domains) was conducted during high nocturnal ozone episode to improve the prediction of the regional ozone concentration in the southeastern coastal area of the Korean peninsula. The analysis was performed by six simulation experiments: (1) without nudging (e.g., CNTL case), (2) with observation nudging (ONE case) to all domains (domain 1∼4), (3) with grid nudging (GNE case) to all domains, (4)∼(6) with grid nudging to domain 1, domain 1∼2 and domain 1∼3, respectively (GNE-1, GNE-2, GNE-3 case). The results for nudging techniques showed that the GNE case was in very good agreement with those observed during all analysis periods (e.g., daytime, nighttime, and total), as compared to the ONE case. In particular, the large effect of grid nudging on the near-surface meteorological factors (temperature, relative humidity, and wind fields) was predicted at the coastline and nearby sea during daytime. The results for application domains showed that the effects of nudging were distinguished between the meteorological factors and between the time periods. When applied grid nudging until subdomain, the improvement effects of temperature and relative humidity had differential tendencies. Temperature was increased for all time, but relative humidity was increased in daytime and was decreased in nighttime. Thus, GNE case showed better result than other cases.