I understand that ERA5 data is not a real representation of any specific “site” but an average value over the grid box that they represent and averaged over a certain time period, e.g. as explained here:
https://confluence.ecmwf.int/display/CKB/Parameters+valid+at+the+specified+time
Firstly, what is the “time average” duration for the ERA5 data e.g. for the 10m V & U -component of wind which can be downloaded from the CDS dataset “reanalysis-era5-single-levels”? Perhaps this is a 6second gust wind speed, or 10min average or 1hour average or something else entirely? Is the same idea also applied to the significant wave height data provided in this dataset e.g. is the wave height from the “Significant height of combined wind waves and swell” also an average significant wave height over a particular time duration (perhaps 1hour as this is the space between data points) or is the maximum value observed in the time period for each data point or again something else entirely?
Secondly, does anyone know of any methods that can be used to estimate/adjust from the spatially averaged data to get back to an estimate of what might be observed at a single point/site within the grid box for wind and waves? I understand observations/measurements of the wind speed from meteorological sites are used to validate/constrain the ERA5 model (or perhaps I’m wrong about this) so presumably there are methods to take site observations and estimate the expected value of the ERA5 data – is there a way to reverse this and get an estimate of the expected site values from the ERA5 data? If there is enough observational data from meteorological sites it might be possible to back-calculate an approximate relationship between site measured data (such as the 10min average wind speed at 10m height or other wind speed definition) and that given in the ERA5 data, does anyone know if this or anything similar has been done?
