Offshore Regional Winds
SeaWind is a collection atmospheric hindcast that have been carried out using the Limited Area models (LAMs) driven by the global reanalysis that provide the initial and boundary conditions. The SeaWind atmospheric fields can be considered proper representations of the atmospheric conditions and provide the offshore winds.
The used LAM model is the WRF model (Weather Research & Forecasting), developed at the National Center For Atmospheric Research (NCAR) in collaboration with other agencies and Universities in the US (Skamarock et al. 2008). The optimal setup of the model and physical parameterization has been selected after several tests to provided an accurate representation of hourly marine-wind variability, especially capturing the extreme-wind events.
SeaWind encompasses several wind reanalysis with different spatial resolutions, lateral boundaries and versions of the model WRF-ARW. The Table below shows all the atmospheric hindcast datasets that are available in IH Data with the main characteristics.
Table 2.1. Main characteristics of the atmospheric hindcast datasets available in IH Data, the coordinates of the geographical coverage are aproximated.
The domains of each hindcast are shown in the Figure 2.1. The source of static data used to get the simulation domains are MODIS Land Cover Type products (Moderate Resolution Imaging Spectroradiometer). The subproduct is the Modified IGBS MODIS 20-category vegetation (land-use) data (https://lpdaac.usgs.gov).
Figure 2.1. Simulation domain showing the topography and coastal boundaries. Dashed lines show the sponge zone where model data is relaxed to the data specified at the boundary.
Figure 2.2. (Top) Temporal correlations and BIAS error of surface wind speed against altimeter observations for the 1992-2015 period. (Bottom) Scatter diagrams, qq-plots and several statistical indices of the buoy measurements (x-axis) and the WRF model results (y-axis). Diamonds represent the quantile values in equally spaced Gumbel scale. Coloured dots are the data pairs of wind speed (m/s), which colour intensity represents the density of data.
A validation of the wind hindcast products against observations (buoy and satellite data) was also obtained for each hindcast datasets. The validation results and more information about SeaWind I and SeaWind II can be found in Menendez et al. (2014).
The validation of SeaWind Cantabrian Sea with a horizontal spatial resolution of 3 km are shown in the Figure 2.2. The validation graphics show the scatter and quantile plot together with several diagnostic statistics of the wind reanalysis against in-situ recorded measurements and a spatial validation results with the Pearson correlation coeficient and BIAS error.