P65 A generalization procedure for Wind Resource Atlas using WRF output

Hahmann, Andrea N., Claire L. Vincent, Jake Badger, and Mark Kelly, DTU Wind Energy, Denmark

Boundary layer winds and other atmospheric parameters from regional reanalyses are often used to estimate wind power resources in regions with limited quality wind observations. Over land, the validation of these simulations against observations is hindered by the representativeness of both the observations, which are influenced by the local terrain and surface roughness, and the simulations, which are very dependent on how the topography and surface roughness are characterized at the scale of the simulations. These issues also arise when the output from WRF is used to drive microscale model simulations. In this instance, one must be careful not to “double count” the effects of terrain speed up and roughness change.

To minimize these problems a method of generalization is introduced and its use demonstrated by several examples. The wind atlases of the North and Baltic Seas and South Africa are used as examples because good observational networks with tall observations are available in these regions. Idealized simulations are also used to demonstrate the response of WRF to a given roughness and to roughness changes.