In 2005 the Energy research Centre of the Netherlands (ECN) published its first version of the Offshore Wind Atlas of the Dutch part of the North Sea [3]. This version has been updated and improved using longer time series and another approach for the calculation of the roughness
...
In 2005 the Energy research Centre of the Netherlands (ECN) published its first version of the Offshore Wind Atlas of the Dutch part of the North Sea [3]. This version has been updated and improved using longer time series and another approach for the calculation of the roughness of the sea surface. In contradiction to other Wind Atlases which are based on measurements [28], use is made of data from the Numerical Weather Prediction model Hirlam. Measurements of wind speeds and directions are only used to validate the Wind Atlas. For the Offshore Wind Atlas, the Hirlam data is interpolated where for the vertically interpolation use is made of the Businger-Dyer profiles in combination with the Monin-Obukhov length [3]. One of the required parameters for the interpolation is the surface roughness. For land, it can be assumed constant while for sea it is variable. In the previous version of the Offshore Wind Atlas, the sea surface roughness has been determined using Charnock’s relation [9], where the so-called Charnock parameter is constant. In the new version, the equation of Hsu is introduced which states that the Charnock parameter is variable and dependent on the wave steepness i.e. the wave height divided by the wave length [19]. Assuming that the North Sea is a shallow sea and using the general wave equation, which relates the sea depth and wave length to the phase velocity of the waves, it was found that the wave steepness can be rewritten in a fraction of the wave height over the wave period multiplied by the square root of the sea depth times the gravitational acceleration. These quantities are derived from measured values which are interpolated to the location of interest. Using this approach, it is tried to improve the prediction of the wind speed distributions for a given location and altitude. Using wind measurements at several locations it was found that adding the wave data to the computations show a small improvement in the estimation of the wind speed distribution compared to the previous version of the Offshore Wind Atlas. For each measurement location and method, a two parameter Weibull distribution has been made, after which a comparison was done between the various shape and scale parameters. Generally, the scale parameter was overestimated by both versions of the Offshore Wind Atlas compared to the measurements. The cause of this behavior might be found in the data used to make the Atlas. The shape parameter is well predicted by the new version of the Offshore Wind Atlas due to the use of wave data. The influence of the wave data is found to be larger for lower altitudes than for higher altitudes. Besides Weibull distributions, also maps with average wind speeds are given by the Offshore Wind Atlas which are compared to older maps.