The main aims of our project were to test the compatibility between the reanalysis data and WAsP and to determine whether or not using the reanalysis data and SRTM maps are suitable to estimate the wind resource of a site as a substitute to extensive sight survey and erecting a meteorological mast.

The first step in evaluating the accuracy of reanalysis data was to establish a correlation between the reanalysis data and some data measured directly on site provided by Scottish Power, via SgurrEnergy. We have established this correlation on 5 sites within the UK (Dun Law, Kentish Flats, Dounreay, Coldham and Beinn Tharsuinn). Comparing monthly these two data sets, we have observed a clear linear correlation with a high factor of determination for both wind speed and wind direction (always greater than 0.7, mainly around 0.9). It seems then that the reanalysis data is in accordance with the measured data of the studied sites.

The next step was then to use these reanalysis data with WAsP in order to evaluate the accuracy of our methodology. We have been able to test our methodology on 2 wind farms (Dun Law and Hagshaw) and on 4 other meteorological sites within the UK (Kentish Flats, Dounray, Coldham and Beinn Tharsuinn). And it appears regarding these results that the reanalysis data can be used with WAsP without many difficulties (after processing), and that these reanalysis data is suitable to estimate the wind resource of a site.

Indeed, the wind resource estimation by WAsP for the studied wind farms stays within 10% of the actual energy produced by the farms when the 27 year average of reanalysis data is used. The wind speed calculated by WAsP is slightly different using measured or reanalysis data (less than 10% different) at the 4 mast sites.

Of course, this methodology has only been tested at a few sites, which isn't enough to establish a general trend or to conclude definitively on the suitability of the reanalysis data to be used to evaluate the wind resource of a site (which is not it's original application, according to Wesley Ebisuzaki: "The NCEP/NCAR Reanalysis is not the appropriate product to use for studying wind energy"). 

We can only conclude that these results seem very promising but the methodology should be tested on other established sites all over the world in order to confirm the first results and to conclude definitively on the accuracy of the methodology. It seems very important to carry out some simulations and comparisons at different places of the globe in order to establish a general application of the methodology.