Calculation of extreme wind atlases using mesoscale modeling

https://energi12.energiforskning.dk/da/project/calculation-extreme-wind-atlases-using-mesoscale-modeling



The objective of this project is to develop new methodologies for extreme wind atlases using mesoscale modeling. Three independent methodologies have been developed. All three methodologies are targeted at confronting and solving the problems and drawbacks in existing methods for extreme wind estimation regarding the use of modeled data (coarse resolution, limited representation of storms) and measurements (short period and technical issues).

The first methodology is called the selective dynamical downscaling method. For a chosen area, we identify the yearly strongest storms through global reanalysis data at each model grid point and run a mesoscale model, here the Weather Research and Forecasting (WRF) model, for all storms identified. Annual maximum winds and corresponding directions from each mesoscale grid point are then collected, post-processed and used in Gumbel distribution to obtain the 50-year wind.

The second methodology is called the statistical-dynamical downscaling method. For a chosen area, the geostrophic winds at a representative grid point from the global reanalysis data are used to obtain the annual maximum winds in 12 sectors for a period of 30 years. This results in 360 extreme geostrophic winds. Each of the 360 winds is used as a stationary forcing in a mesoscale model, here KAMM. For each mesoscale grid point the annual maximum winds are post-processed and used to a Gumbel fit to obtain the 50-year wind.

For the above two methods, the post-processing is an essential part. It calculates the speedup effects using a linear computation model (LINCOM) and corrects the winds from the mesoscale modeling to a standard condition, i.e. 10 m above a homogeneous surface with a roughness length 5 cm. Winds of the standard condition can then be put into a microscale model to resolve the local terrain and roughness effects around particular turbine sites. By converting both the measured and modeled winds to the same surface conditions through the post-processing procedure, it becomes possible to validate the measurements more reasonably.

The third method is called the spectral correction method. It is targeted at improving the general smoothing effect of mesoscale models for extreme wind application. Through wind time series of one year or more, or a spectral model if there are no measurements, we can add the wind variability to the mesoscale modeled time series, thus obtaining improved extreme wind estimation.

The first method has been applied to three areas with different degrees of terrain complexity, Denmark (flat), Gulf of Suez (medium complex) and Navara region in Spain (highly complex), and to places with different extreme wind mechanisms (the synoptic lows and the channeling winds). The second method has been applied to Denmark and Gulf of Suez. The two methods give quite consistent extreme wind atlases which agree well in general with the measurements from several sites; the first method has given better estimates for Gulf of Suez. The first method is more cost effective for larger areas. The third method can generally be applied to any mesoscale modeled wind time series.

Even though the extreme wind atlases obtained with the new methodologies are quite satisfactory for non-complex terrains, challenges remain in (1) very complex terrains, e.g. mountains with steep cliffs, (2) coastal areas where land and water transit, (3) offshore where the unique wave dynamics during storm conditions have not been taken into account in the atmospheric modeling of the current methods.

Three independent methodologies have been developed: The storm episode method; the spectral correction method and the statistical-dynamical downscaling method. All three methodologies are targeted at confronting and solving the problems and drawbacks in existing methods for extreme wind estimation regarding the use of modeled data (coarse resolution, limited representation of storms) and measurements (short period and technical issues).

Formålet er at udvikle bedre oversigter over ekstreme vindforhold, der er nyttig for vindmøllefabrikanter og ved placeringen af vindmøller. Det væsentlige metode i projektet er at benytte mesoskala modeller til fastlæggelse af ekstreme vinde og importere data ind i WAsP softwaren.