To validate the results of MesoMap, data from diverse meteorological towers distributed all over Spain was used. All the towers selected were higher than 40 m. The above mentioned information was processed in terms of mean annual wind speed (m/s), total annual data availability (%), measurement period length (months), sensor height (m) and coordinates of the measurement sites. The data form all 47 stations were extrapolated to the agreed level of 80 m, in order to be compared with the results from modelling.
The following graphs, for its simplicity, allow drawing some generic conclusions about the results. First of all, model behaviour is described with respect to the measurements:
In general, atmospheric simulation with the resolution used in this study allows locating and distinguishing the sites of interesting wind resource, as shown by measurement trend tracking.
The following graph, which represents the uncertainty of the study (modelling - real value extrapolated to 80 m above ground level) depending on wind speed, illustrates the behaviour and limitations of the modelling techniques:
It can be seen that the model tends to overestimate low wind speeds and underestimate high wind speeds. This is caused by an unavoidable technical limitation, since one of the effects of any modelling system is its smoothing of the fields describing any variable, when representing in one only node all the values of its environment . The model MASS divides the atmosphere in a three-dimensional grid. In every node of this grid, variables take an average value that is representative from a cube around this node, which leads to information loss in the end. This effect is smaller as microscale resolution is higher, but it exists even in resolutions as high as used in this study (a 100 m grid).
To summarise, the following conclusions are drawn: