| dc.description.abstract | The variations in surface temperature drive weather dynamics in the atmospheric boundary layer. Consequently, accurate prediction of surface temperature is paramount to understanding the weather. The present study attempts to predict surface temperature using the Weather Research and Forecast (WRF) model. November 2017 simulation was used to investigate two sources of Lateral Boundary Conditions (LBCs), that is, European Centre for Medium-Range Weather Forecast Re-Analysis (ERA-interim) and National Centre Environmental Prediction-Final analysis (NCEP-FNL). A further investigation was conducted on the sensitivity of the simulated surface air temperature to Planetary Boundary Layer (PBL) parameterization schemes. These include two nonlocal schemes, that is, Yonsei University, (YSU), Asymmetrical Convective Model version 2, (ACM2) and three local schemes, that is, Quasi-Normal-Scale Elimination, (QNSE), Mellor-Yamada Nakanishi Niino version 2 (MYNN2), Mellor-Yamada-Janjic (MYJ) and their effects on the simulated surface air temperature. The results show that the ERA-interim provided boundary conditions that resulted into a higher skill in simulating temperatures compared to the NCEP-FNL. Subsequently, simulations done with ERA-interim LBCs were used to investigate the sensitivity of the simulated surface temperature to PBL parameterization schemes, and the results show that MYNN2, MYJ, YSU and ACM2 PBL schemes captured better characteristic variations of the surface temperature. However, these PBL schemes tended to produce their peak values 6-hours earlier than the observations. Although the QNSE scheme overestimated temperature compared to the surface observations, it produced a strong correlation coefficient compared to the rest of the schemes, meaning that the simulated surface temperature followed a similar pattern to the observations. The results generally showed that the local scheme, MYJ provided better simulation of surface temperature compared to the other schemes and is recommended for use in operational weather prediction. | |
