[세미나] 유영희 박사님

The importance of clouds in photochemistry and air quality modeling: A quantitative assessment using satellite cloud retrievals

The importance of clouds in photochemistry and air quality modeling: A quantitative assessment using satellite cloud retrievals

Accurate cloud representation in air quality models plays a key role in photochemical production of secondary pollutants such as ozone (O3) and their prediction skills. It is very important, therefore, to understand and quantify errors in O3 prediction associated with errors in clouds. To quantify the errors in O3 due to inaccurate cloud predictions, two approaches are introduced. First, satellite cloud products are employed in the WRF-Chem model to constrain radiation fields for photochemistry, and the effects of satellite cloud-constrained radiation on summertime O3 formation over CONUS are evaluated. O3 simulations for 2013 summer are carried out using reanalysis data, and the results show that the average difference in summer time surface O3 concentrations derived from the modeled clouds and satellite clouds ranges from 1 to 5 ppb for maximum daily 8-h average O3 (MDA8 O3) over CONUS. The cloud bias accounts for up to ~40% of the total MDA8 O3 bias under cloudy conditions. It is found that O3 concentrations are sensitive to cloud errors mainly through the calculation of photolysis rates (for ~80%), and to a lesser extent to light-dependent biogenic volatile organic compounds emissions. The sensitivity of O3 to satellite-based cloud corrections is about 2 times larger in VOC-limited than NOX-limited regimes. These results highlight that the benefits of accurate predictions of cloudiness would be significant in VOC-limited regions which are typical of urban areas. Second, the benefits of cloud assimilation on O3 production and forecast skills are examined using two different initial meteorological conditions, i.e., the Global Forecast System (GFS) and Rapid Refresh (RAP) in WRF- Chem simulations during 2016 summer. RAP assimilates METAR observations and satellite-retrieved cloud top pressure to modify background hydrometeor fields. The RAP simulations are found to perform better for the next-day O3 forecasts than the GFS simulations as O3 in GFS simulations is generally overpredicted. The mean bias error of MDA8 O3 over CONUS is smaller by ~3 ppb under cloudy conditions in RAP simulations than in GFS simulations. The differences in predictability of cloud cover between RAP and GFS simulations are analyzed, and the results show that the predictability of cloud cover is the highest for 1-h forecasts using RAP and gradually decreases with the forecast length. Among other meteorological parameters, boundary layer height is found to play an important role in O3 production and thus O3 forecast skill.