Posters - WindEurope Technology Workshop 2022
Resource Assessment & Analysis of Operating Wind Farms 2022
23-24 June • Brussels


Come meet the poster presenters to ask them questions and discuss their work

Check the programme for our poster viewing moments. For more details on each poster, click on the poster titles to read the abstract.

PO078: Wind-validation: a free and open-source python package to simplify validation

Rogier Floors, Senior Researcher, DTU Wind Energy


­­­­The wind energy industry is using a plethora of models to do wind resource assessments and nearly all these models are regularly validated with measurements. These validations are often one-off activities and a unified way to validate models does not exist. As part of the WindSider project, DTU developed a free and open-source python package that can do validation for three different wind resource data types that are commonly used: wind speed and direction time series, histograms, and Weibull distributions. The inputs and outputs of the package are xarray datasets, which allow storage of multi-dimensional, labelled data. They can be made using the accompanying package windkit. A key benefit of this package is the ability to add metadata to objects, which makes datasets findable, accessible, interoperable, and reusable (FAIR). In this presentation we show the basic structure of the package: it allows validation of wind climates based on statistics and metrics. Statistics are single variables that can be calculated from the observations or the model, like the mean wind speed. A metric compares a model with observations, for example the mean absolute error or correlation coefficient. We show three examples where the package has been used. This includes a validation of reanalysis, mesoscale and microscale wind climates, a validation of the WAsP profile model and a validation of a dataset from scanning lidars. Because many of the data are open source and integrated with the wind validation package, the user can easily validate their own model as well. A simple reporting engine is included in the package, which generates a report that opens automatically in the browser from within python. It contains aggregated metrics and statistics for an overview, a map of the included sites and detailed plots that allow interpretation of the results. It also contains a reporting interface for comparing several models or different versions of models with observations. These reporting tools can form the basis for benchmark exercises and dialogue about wind resource model performance in general.