|
1. anemos wind atlas
The anemos wind
atlas for Germany represents a database containing long-term time series for
all relevant atmospheric parameters. anemosuses
the mesoscale and numeric weather prediction model system WRF3.7.1 for
downscaling the reanalysis data.
First, MERRA-2 reanalysis data are used as input and driving data. Before the
WRF main simulation starts, it is necessary to proof the model approach and parametrisations (e.g.
planetary boundary scheme, surface pattern, radiation scheme) and optimize
the model for the relevant parameters (wind speed and direction).
2. Wind atlas verification
The basic assignment after the simulation with WRF is the intensive
verification of the anemos wind atlas for Germany with the aid
of numerous internal measurements. On the one hand the verification of the
wind atlas gives a dimension of prediction benignity and quality of the
simulation output. On the other hand a systematic bias can be corrected in
the last step; the Remodelling, to improve the
quality of the anemos wind atlas. The verification
analyses statistic parameters as mean value, Pearson correlation, bias, RMSE
and extreme values (QQ-distribution). Furthermore anemos verifies
the vertical profile, diurnal and annual cycle and frequency distribution
with weibull parameters.
3. Remodelling improvements
The Remodelling is
an optimization process for the wind speed. Scaling factors are developed
with the help of numerous internal measurements and the subgrid topographic
information (orography, roughness, etc.).
Finally the Remodelling improves
the statistic parameters as well as the frequency distribution, diurnal and
annual cycle.
4. References
[1] Skittides C.
and Früh W.-G., 2015, A new
Measure-Correlate-Predict Wind Resource Prediction method, ICREPQ'15
[2] Staffell I.
and Pfenninger S.,
2016, Using bias-corrected reanalysis to simulate current and future wind
power output, Energy 114, 1224-1239.
|