Creating and validating high-resolution wind atlases

<p>Wind-atlases are vital for planning of the massive transition towards renewable energy that is required over the next decades. Reanalysis products like ERA5 are frequently used to obtain winds at hub heights, but coarse reanalyses products often lack crucial details that are needed to represent the microscale flow. To circumvent this aspect, wind-atlases use a procedure of generalizing and downscaling mesoscale model outputs to predict the flow down to the microscale (~250 m grid spacing). The methods to do this are based on the geostrophic drag law and the logarithmic wind profile, which both need information from the mesoscale model to represent for example the wind profile in each wind-direction sector. However, one also needs high-resolution roughness and elevation maps to represent microscale speed-up effects. Combining these large-scale and microscale effects is notoriously difficult</p><p>Due to lack of measurements it has been difficult to validate the accuracy of these wind atlases at the hub heights of modern turbines. In this presentation we briefly discuss recent improvements in the model chain related to stability and surface roughness modelling, but mostly focus on a validation of these new wind atlases at more than 60 tall masts around the world. Three python packages have made it much easier to generate and validate these wind atlasses. The first package contains wind-related data structures including geospatial information that make your data self-explanatory. A second python package to make wind validations easier to perform is also introduced. It contains a work flow to validate data and create simple reports to analyze them. Finally, flow modelling is done using PyWAsP, which contains submodules for roughness, stability and wake-modelling.</p>