Validating the Water Vapor Variance Similarity Relationship in the Interfacial Layer Using Observations and Large‐Eddy Simulations

In previous work, the similarity relationship for the water vapor variance in the interfacial layer (IL) at the top of the convective boundary layer (CBL) was proposed to be proportional to the convective velocity scale and the gradients of the water vapor mixing ratio and the Brunt‐Vaisala frequency in the entrainment zone. In the presence of wind shear in the IL, the similarity relationship was hypothesized to also include a dependence on the gradient Richardson number. Simultaneous measurements of the surface buoyancy flux, wind‐shear profiles from a radar wind profiler, water vapor mixing ratio and temperature measurements and their gradients from a Raman lidar provide a unique opportunity to thoroughly examine the function used in defining the variance and validate it. These observations were made over the Atmospheric Radiation Measurement Southern Great Plains site. We identified 19 cases from 2016 during which the CBL was quasi‐stationary and well mixed for at least 2 hr in the afternoon. Furthermore, we simulated the CBL using a large‐eddy simulation (LES) model for these cases and derived the water vapor variance and other profiles to test the similarity function. Utilizing this unique combination of observations and LES, we demonstrate that the water vapor variance in the IL has little‐to‐no dependence on wind shear. Furthermore, we demonstrate that the predicted variance using the original similarity function matches the observed and LES‐modeled variance very well, with linear correlations between the two variances of 0.82 and 0.95, respectively. Numerical weather prediction and global circulation models need to be able to predict the variance in water vapor in the atmosphere, as this is an important signature of turbulent mixing. However, the variance is something that is not directly resolved by the model and must be approximated using variables that the models actually resolve such as gradients in water vapor, temperature, and wind. This study evaluates a commonly used approximation approach, illustrates its shortcomings, and suggests how the approximation can be improved. A similarity relationship is used to predict water vapor variance in the interfacial layer from other variablesWind shear is shown to be unimportant for the prediction of water vapor variance in the interfacial layerThere is very high correlation between the true and predicted variance in both the observations and LES data