Numerical Simulation of Horizontal Convective Rolls Over a Tropical Coastal Site Using WRF: Sensitivity to Land Surface Physics

Horizontal convective rolls (HCR) are sub-mesoscale motions that influence the transport of heat, momentum, and pollutants within the boundary layer. In this work, a high-resolution (0.666 km) Weather Research and Forecasting (WRF) model is employed to simulate the structure of the HCRs over Kalpakkam along the southeast coast of India. The sensitivity of HCR simulation to model land surface physics is studied with two land surface models (LSM), (i) Noah and (ii) Noah multi-parameterization (NMP), for three selected days (15 April 2013, 07 May 2015, 28 March 2018) during summer synoptic conditions. On all three selected days, the boundary layer rolls formed over a period of about 2–3 h in the morning under moderate winds (4–5.0 m/s), moderate vertical wind shear (2.4–3.5 m/s) in the lower atmosphere, and slightly unstable conditions [gradient Richardson number (Ri_G) −4.5 to −5.0] in both simulations and observations, indicating that thermal instability is the chief mechanism in their development. Simulated mean surface meteorological parameters by NMP were found to be in better agreement with observations than Noah. Results suggest that the LSMs mainly affected the simulated turbulent roll structure in terms of updraft cores and their horizontal and vertical extent by variation in simulated surface energy fluxes, boundary layer structure, wind shear, and stability. The structure of simulated HCRs is better represented by NMP due to the improvements in the flux distribution and surface properties. Simulations using the FLEXPART dispersion model for a hypothetical case of tracer release indicated an uneven spatial concentration pattern due to upward and downward motions in the region of HCRs. The stronger winds and stronger flow convergence in Noah and higher heat flux and more unstable conditions in NMP led to differences in the simulated tracer concentrations in the two cases.