Improving the Water Vapor Variance Similarity Relationship in the Interfacial Layer Using Raman Lidar and Radar Wind Profiler Observations with LES

Similarity relationships are used to predict higher-order moments (e.g., the variance and flux of water vapor, sensible heat, and momentum) from variables that numerical weather prediction and climate models actually predict (e.g., mean profiles and gradients). Advances in remote sensors, especially in lidar technology, have allowed datasets to be collected that can be used to evaluate the accuracy of these similarity relationships.

The water vapor variance similarity relationship in the interfacial layer (IL) on the top of the convective boundary layer (CBL) has been proposed to be proportional to the convective velocity scale (w*), water vapor gradient at IL, Brunt-Vaisala frequency at IL, and the wind shear across the IL. The ARM Southern Great Plains site has the capability measure all of these terms using a combination of a surface energy balance station, Raman lidar, and radar wind profiler. In particular, this Raman lidar is able to measure both water vapor at sufficiently high temporal and vertical resolution to derive both the profile of variance and vertical gradient, as well as temperature profiles with enough resolution to derive the vertical gradient.

Nineteen cases were identified in 2016 where the CBL was quasi-stationary and could be analyzed to derive the variance and gradient profiles to analyze the similarity relationship. These cases were also simulated using the microHH large eddy simulation (LES) model. The LES model was driven with the operational variational analysis product routinely produced at the SGP by the ARM program. The combination of the LES and observational datasets demonstrate that the water vapor variance in the IL has no dependence on wind shear. Furthermore, the variance predicted using a modified version of the similarity function matches the observed and LES-modeled variance very well, with correlations between the two variances of 0.82 and 0.95 respectively.