Sensitivity of Planetary Boundary Layer Scheme in WRF-Chem Model for Predicting PM10 Concentration (Case study: Jakarta)

Abstract

Weather Forecasting-Chemistry (WRF-Chem) is an air pollution model that combines meteorological factors and atmospheric chemistry simultaneously (online coupled). Many schemes selected on WRF-Chem according to boundary layer conditions of the study area. The objectives of this study were to evaluate WRF-Chem model and analyze the sensitivity of Yonsei University (YSU) scheme and Mellor Yamada Janjic (MYJ) scheme as physical boundary layer parameters, especially for Jakarta region. Particulate matter 10 micrometers or less in diameter (PM10) selected as the air pollutant parameter. The model outputs compared to observational data in five locations in Jakarta. The regression analysis between air temperature data shows that the values of R2 are 0.5 - 0.6 in February and 0.5 - 0.9 in August. Another result, the wind velocity from WRF-Chem overestimates but the hourly averages have a similar diurnal pattern with the observation. The PM10 concentration model output is underestimated which is caused by local emissions that are not accommodated in the model. The sensitivity analysis shows a higher sensitivity index for MYJ than YSU scheme, especially in August, so the model outcome of the MYJ scheme has a good response to the observations in Jakarta. It is seen that the extreme concentration of PM10 produced through the WRF-Chem with both schemes in February and August have the same time in morning. The model output of WRF-Chem through both schemes also demonstrate the minimum concentration of PM10 that follows during afternoon. The Planetary Boundary Layer (PBL) height analysis shows that in morning until noon, the PBL height with MYJ scheme is lower than YSU. It is different for other hours, that from the afternoon through the night until next morning, the height of PBL with MYJ scheme is thicker than YSU. At the point when the PBL is high, the pollutants assorted in a more volume so the estimation of the pollutant concentration estimated on the surface will be reduced.