Added value of high resolution forecast models in the assessment of UK offshore wind resource

摘要

In 2014 the Met Office were awarded a contract by The Crown Estate to produce a new UK wide offshore wind dataset at a height of 110m. This has been created using two forecast datasets produced using the Met Office's numerical weather prediction (NWP) model, the Unified Model. One of the datasets is a 4 year archive of high resolution (1.5km) operational forecast data; the other is 30 years of a 4.4km resolution hindcast dataset, produced specifically to be a long-term, high resolution and consistent dataset. The matching 4 years of data between the two datasets were used to calculate linear regression coefficients based on twelve 30° wind direction sectors. These were then applied to the full 30 years of 4.4km data to produce a dataset which utilised the high resolution, optimal configuration and data assimilation of the operational 1.5km archive as well as benefitting from the long term variability of the 30 year hindcast dataset. The approach has been verified against met mast measurements and an improvement is seen above just using the 30 year dataset, especially in the absolute biases. The increased resolution in the 1.5km model archive results in a better resolved coastline which enables the new UK wide offshore wind dataset to have more detail over these coastal areas. Using two pre-existing datasets in this way was a computationally efficient method and shows the advantages that are possible to gain through combining different forecast datasets, including operational models.