The structure of the atmospheric boundary layer (ABL), particularly the convective boundary layer (CBL) that develops in the ABL during the daytime, is strongly driven by the transport of momentum, heat, and water vapor on the surface. The roughness length for momentum (Z0m) and the roughness length for heat and water vapor (Z0ℎ) are key parameters in the equations describing surface heat budget and transport. Previous studies have indicated that settings of the roughness length influence surface air temperature and flux calculations. However, numerical models for land surface processes often assume Z0m = Z0ℎ or rely on insufficient parameterizations. While Z0m and Z0ℎ are parameters with physically different characteristics, the impact of their difference on the development of the CBL and the time evolution of surface meteorological parameters remains unclear. The purpose of this study is to evaluate the sensitivity of surface meteorological parameters to variations in Z0m and Z0ℎ during the development of the CBL.

In this study, idealized experiments assuming an urban surface were conducted using the Cloud Resolving Storm Simulator (CReSS; Tsuboki, 2023). Additionally, sensitivity experiments were conducted by varying three parameters (evaporation efficiency β ,Z0m and Z0ℎ) across five different settings for each. The idealized experiments confirmed that surface meteorological parameters change significantly with increased roughness. Sensible heat flux decreased due to reductions in surface wind speed and the temperature difference between the surface and the atmosphere. Latent heat flux increased in spite of the reduction in surface wind speed and the water vapor mixing ratio difference.

In the sensitivity experiments, regarding the sensitivity of sensible heat flux, the contribution of the bulk coefficient for heat and water vapor was found to be significant. The reason for the low sensitivity of sensible heat flux to changes in Z0m was that the increase in Ch counteracted the decrease in surface wind speed caused by the increase in Z0m. The reason for the low sensitivity of sensible heat flux to changes in Z0ℎ was that the increase in Ch counteracted the decrease in temperature difference caused by the increase in Z0ℎ.

Similarly, regarding the sensitivity of latent heat flux, the contribution of the bulk coefficient for heat and water vapor was found to be significant. The reason for the low sensitivity of latent heat flux to changes in Z0m was that the increase in Ch counteracted the reductions in surface wind speed caused by the increase in Z0h. The reason latent heat flux increased with the increase in 𝑍0ℎ was that there were no other factors to cancel out the increase in Ch caused by the increase in Z0ℎ.

It was found that urban model calculations using conventional settings for Z0ℎ are inappropriate because they result in excessive latent heat flux; thus, the setting for Z0ℎ in urban areas must be a significantly smaller value. However, simply reducing Z0ℎ causes an undesirable decrease in sensible heat flux. To address this, the urban heat budget can be reproduced more accurately by defining two distinct roughness lengths: one for heat and another for water vapor. By using independent bulk coefficients, sensible and latent heat fluxes can be controlled individually.

Based on these results, it is concluded that the setting of surface roughness, specifically the roughness for heat and water vapor, plays an extremely important role in determining the amount of energy supply (flux) to the CBL and the associated surface meteorological parameters.