The Atmospheric Boundary Layer in Numerical Weather Prediction
(part of the Hans-Ertel-Centre for Weather Research)
The atmospheric boundary layer (ABL) plays an important role in numerical weather prediction and climate simulations. Its structure and evolution have a strong impact on near-surface weather and climate. ABL processes, such as turbulence and coherent motions, for example, contribute to the formation and development of clouds and thunderstorms. They also largely control the exchange of momentum, heat, water and other constituents between the land surface and the free atmosphere.
Representing the stable ABL in weather and climate models, in particular, poses a great challenge. In the stable ABL, turbulence is weak and intermittent. Other processes such as radiation and small-scale coherent motions become more important in determining the characteristics of the ABL. The influence of these other processes on the ABL is poorly known and currently not, or only crudly, represented in weather and climate models.
Our group studies the atmospheric boundary layer over complex terrain, with the goal of improving our understanding of its structure and evolution and our ability to represent and predict it in weather and climate models. Research questions that our group are addressing include:
- How can we best represent the impact of small-scale clouds and their organization on the boundary layer in atmospheric models?
- What are the dominant small-scale motions in the ABL over flat and heterogeneous terrain? By which processes are they caused and how can the impact of these processes be represented in atmospheric models?
- What is the impact of small-scale orography on the ABL and the lower atmosphere? How do topographically induced motions such as, for example, gravity waves, slope and valley winds, impact the exchange of momentum, heat, and mass in the mountain boundary layer?
These questions are addressed by combining theory, observations, and atmospheric models of various complexity. Our main research tools are large-eddy simulation (LES; link) and Doppler Lidars. LES with a resolution of O(1-100 m) allows for the explicit representation of the largest turbulent eddies and small-scale coherent motions in the atmospheric boundary layer under stable and convective conditions, respectively. Doppler Lidars enable the measurement of coherent motions and turbulence statistics in the ABL. Based on these simulations and theoretical developments, parameterization ideas are developed and tested in single-column and numerical weather prediction models.
Figure 1: Snap shots of the cloud liquid water path from three different LES of shallow cumulus convection over the ocean without mesoscale organization (CTRL) and mesoscale organization of the cloud field (STD and MST).
Figure 2: Vertical cross section of velocity perturbation from a direct numerical simulation (DNS) of a stratified boundary layer flow over gently sloping terrain.
Figure 3: Moisture flux for a shallow convection case for a standard parameterization scheme (ICON-NWP-TURB+CONV), our new unified cloud-turbulence scheme (ICON-2TE), and a reference LES.
Field Experiment on submesoscale spatio-temporal variability in Lindenberg (FESSTVaL)
Our group is part of the FESSTVaL field campaign in summer 2020 that focuses on sub-mesoscale variability. Thereby, our aim is to
- Document the turbulent characteristics of typical summertime boundary layers (clear, cloud-topped, with cold pools)
• diurnal evolution, convective and stable boundary layer
• day-to-day variability (dry, cloudy, with cold pools)
- Identify the dominant scales of turbulent & submeso motions, the processes promoting these motions, and how well they can be identified
- Evaluate boundary layer parameterizations (typical summertime ABL regimes and transition between the regimes)
To reach this aim we will employ a small network of Doppler wind LiDARs of type HALO Streamline that will be set up in the FESSTVaL domain. We plan to combine different scan strategies, including VAD scans, RHI and PPI scans as well as vertical stares to get an optimal information about the turbulence and sub-mesoscale motions, as well as their spatial distribution.
The field campaign will be accompanied by Large Eddy Simulations (LES). This will first of all help us to determine an optimal positioning and scan strategy for the LiDARs. Second, it will yield a 3D gridded dataset in addition to the measurements.
Moreover, routine simulations with a Single Column Model configuration of the ICON (Icosahedral – Nonhydrostatic) model will be run to document and evaluate the performance of different boundary-layer parameterization schemes.