Has anyone had a problem with significant wave height (swh) and mean wave period (mwp) from ERA-Interim not matching very well with either a time series of measurements or a different model? Or does anyone know if the formulas in the model (the way they made it) have changed from one period to another? (In my case they seem to correlate better/worse at either side of 1991).
I haven't looked into ERA-Interim wave data a huge amount but I know around 1991 it started to assimilate ocean wave data from satellite altimeters (ERS-1). See the first figure here https://climatedataguide.ucar.edu/climate-data/era-interim
I remember Jean-Raymond Bidlot who runs WAM at ECWMF showing me so experiments with and without the assimilation into the wave. I think the Hs differed by ~1m in large events so not insignificant.
This doesn't answer your question but I can't comment; It is worth noting that there are several changes in SST packages which drive the models. The most significant is the difference from before December 2001 and after January 2002 - it could be that there is a difference from before 1991 and after for similar reasons.
It might also be worth checking if SWH or MWP are data packages used to drive the model or if they are produced at the output.
You can find model formulas here: https://www.ecmwf.int/sites/default/files/elibrary/2011/8174-era-interim-archive-version-20.pdf
and the reanalysis process is described here; http://onlinelibrary.wiley.com/doi/10.1002/qj.828/pdf
Hope this helps! :)