Can machine learning improve the model representation of turbulent kinetic energy dissipation rate in the boundary layer for complex terrain?

• Main Authors: N. Bodini, J. K. Lundquist, M. Optis
• Format: Article
• Language: English
• Published: Copernicus Publications 2020-09-01
• Series: Geoscientific Model Development
• Online Access: https://gmd.copernicus.org/articles/13/4271/2020/gmd-13-4271-2020.pdf
• ISSN: 1991-959X; 1991-9603

<p>Current turbulence parameterizations in numerical weather prediction models at the mesoscale assume a local equilibrium between production and dissipation of turbulence. As this assumption does not hold at fine horizontal resolutions, improved ways to represent turbulent kinetic energy (TKE) dissipation rate (<span class="inline-formula"><i>ϵ</i></span>) are needed. Here, we use a 6-week data set of turbulence measurements from 184 sonic anemometers in complex terrain at the Perdigão field campaign to suggest improved representations of dissipation rate. First, we demonstrate that the widely used Mellor, Yamada, Nakanishi, and Niino (MYNN) parameterization of TKE dissipation rate leads to a large inaccuracy and bias in the representation of <span class="inline-formula"><i>ϵ</i></span>. Next, we assess the potential of machine-learning techniques to predict TKE dissipation rate from a set of atmospheric and terrain-related features. We train and test several machine-learning algorithms using the data at Perdigão, and we find that the models eliminate the bias MYNN currently shows in representing <span class="inline-formula"><i>ϵ</i></span>, while also reducing the average error by up to almost 40&thinsp;%. Of all the variables included in the algorithms, TKE is the variable responsible for most of the variability of <span class="inline-formula"><i>ϵ</i></span>, and a strong positive correlation exists between the two. These results suggest further consideration of machine-learning techniques to enhance parameterizations of turbulence in numerical weather prediction models.</p>