Assessment of WRF Surface Layer Formulations Over a Complex Terrain

The accurate and efficient estimation of surface turbulent fluxes is crucial to predict an adequate atmospheric evolution by atmospheric models. The Monin-Obukhov similarity theory, which is used to compute these fluxes in numerical models, utilizes the empirical stability correction functions. In the present study, impact of various functional forms of similarity functions on the computation of the surface fluxes under both unstable and stable stratification is analyzed. In addition, this study compares two surface layer parametrization schemes in the Weather Research and Forecasting model over Ranchi (India). The model is run with three nested domains at a high resolution (1 Km) for `five' 4-day periods covering 15 days of Premonsoon season. The two surface layer schemes chosen for the analysis includes MM5 surface layer scheme having Businger-Dyer similarity functions, and revised MM5 scheme utilizing the functions those are valid for full ranges of atmospheric stabilities. The five planetary boundary layer (PBL) schemes are selected to assess the influence of the surface layer schemes on the structure of the boundary layer. The schemes are- Asymmetric Convective Model Version 2 (ACM2), Bougeault-Lacarrere (Boulac), Medium Range Forecast (MRF), Mellor-Yamada-Nakanishi-Niino (MYNN), and Yonsei University (YSU) PBL schemes. The impact of surface layer parametrizations on the near surface diagnostic variables is analyzed and results are compared with the observations.The bias in the 2 m temperature (T2) and 10 m wind speed (U) across the PBL schemes is very small and each PBL scheme is able to reproduce the diurnal variation of T2 irrespective of the surface layer scheme used for the simulations. A relatively higher value nocturnal T2 is predicted with the revised MM5 surface layer scheme as compared to that obtained with the old MM5 scheme, while both the surface layer schemes reproduce almost similar T2 during convective conditions. However, compare to the observations, each PBL scheme is found to overestimate the nocturnal T2 and U with both the revised and old MM5 surface layer schemes during stable conditions.The study suggests that the revised MM5 surface layer formulation is not able to reduce the bias in surface layer variables significantly over tropics and reasons for this are critically analyzed.