European Commission logo
English English
CORDIS - EU research results
Content archived on 2024-05-28

Vertical Structure of the Atmospheric Boundary Layer and Applications

Final Report Summary - VSABLA (Vertical Structure of the Atmospheric Boundary Layer and Applications)

The three dimensional wind fields in the atmospheric boundary layer (ABL) are among the major drivers of local weather, climate, air pollution, and renewable energy sources. Therefore, both observations and modeling studies of vertical profiles of the wind are highly needed. Moreover, the complex turbulent processes in the atmosphere require studies of turbulence parameters, the height and structure of the atmospheric boundary layer. Thus, the studies within VSABLA and the related Danish project tall wind aim to define a set of observations and modeling tools needed to improve the wind resource assessment as well as the design of wind generators, tall buildings, other structures, the performance air pollution models and other applications.

The results of VSABLA are published in journals and at scientific and at applied conferences and in this way disseminated and made available for application. VSABLA is also directly linked to wind industry through wind energy industry partners within the Tall Wind project.

The main scientific goals of the VSABLA project were to assess the current ability of meso-scale models and ABL parameterisations to model the wind profile within the entire ABL based on new technology high resolution in time and space measurements (performed at RISOE DTU within the tall wind project); and also to develop new parameterisations.

In line with this, wind speed and direction data from one year measurements at the western coast of Denmark using a pulsed wind-lidar (wind cube 70) were compared with model simulations carried out with the weather research and forecasting (WRF) advanced research WRF (ARW) model of national center for atmospheric research (NCAR), USA on six-kilometer resolution grid (Gryning et al, ICEM2011). It was found that the model underestimates the mean wind speed within 3.3 % at 100 m and within 4.5 % at all heights up to 600 m. Further analysis is being performed on comparisons of modeled and observed momentum flux and roughness parameters (Floors et al, EMS2011). The resolution in height of the data is 50 m in the range from 100 to 600 meters and the resolution in time is ten minutes.

In addition, based on the analysis of these long-term data, a new parameterisation for the shape parameter in the Weibull distribution was suggested and tested (Gryning et al, ICEM2011).

Different physical options, in particular ABL schemes of the WRF ARW model, as well as runs in forecast and analysis mode were evaluated against the measurements for two-week periods (Batchvarova et al, EWEA2011 and Floors et al., ASR, 2011). Generally, the model under predicts the wind speed at all heights above 60 m. Classifying the wind profiles according to thermodynamic stability in the surface layer and normalising them with the measured friction velocity, reveals that the model performs best at neutral conditions, slightly overestimates the wind profile under unstable conditions and underestimates at stable conditions.

The novelty of the results consists of the high resolution of the data in space and time (50 m vertical resolution and ten minutes intervals) within the ABL, provided by the pulsed lidar; the evaluation of model predictions and further development of parameterisations.

Within VSABLA, a novel theory of model evaluation or model-to-data comparison is developed by suggesting a method that combines model prediction and variability of the meteorological parameters, estimated from the measurements (Batchvarova and Gryning, 2010, Batchvarova et al., 2010, Gryning et al., 2011). The variability for wind speed depends on the mean wind speed and the averaging period of the measurements. In case the data fall within the variability interval set along with the model results, the model is performing well and can in principle not be improved. The method was applied in a model comparison exercise performed within COST728, based on data from the observatory of the German weather service at Lindenberg.