Methodology III: GWS® MESO by Al-Pro

GWS® MESO

Model validation	The results of the GWS® MESO simulations are validated with wind data from meteorological stations. The wind data are weighted with respect to its representativeness for the model area and with respect to its quality and the length of the measurement. GWS® MESO model results are then corrected and optimized using the weighted wind data information. Al-Pro has completed a study [21] on the quality of GWS® MESO data based on the wind data of approximately 340 European met stations. The validation check of the mean wind speed was based on reference data from Deutscher Wetterdienst [10], European Wind Atlas [11], the Wind Atlas for the Central European Countries [12] and on data from Wetteronline GmbH [13]. The majority of the reference stations are located at airports. Station positions were also checked.
				The comparison of the GWS® MESO data with the validation data at a reference height of 10 m above ground level shows a deviation of ± 0.5 [m/s] for 52 % of the GWS® samples. A total of 29 % of the calculated GWS® wind speeds are within 1 [m/s] and 10 % are within 1.5 [m/s]. An additional 9 % of calculated GWS® wind speed deviates more than 1.5 [m/s] from the weather station data. Click here to download the study.

Boundary zone	Due to boundary effects of the GWS® MESO computations, a boundary zone with wind values of lower accuracy has to be introduced in the GWS® MESO maps. The boundary zone is displayed as an area with pale colours on the periphery of the map. As a rule, each new GWS® MESO map has such a boundary zone. While continuing with the computation of adjacent maps, the wind velocity values of the boundary zone can be corrected. To do so, 8 maps surrounding one central map have to be computed. The correction of the wind values of the central map’s boundary zone is accomplished by averaging the wind values of the overlapping boundary zones of all 9 maps.
Limitations	GWS® MESO wind maps give an overview of the wind quality which can be expected in the examined area. The data for each computed cell is the climatical average of the area covered by that cell, in this case 2 x 2 km. Wind conditions at a specific location within this area may differ from this average. Especially in complex terrain situations where differences from the averages can be expected - on hilltops and ridges wind speeds might be higher and in valleys lower than the average. Due to the nature of GWS® MESO modeling it is not always possible to completely reproduce local wind conditions. Discrepancies to real wind values can be expected in regard to wind speed and wind direction. Local wind systems for instance, like land and sea breezes, cannot be exactly modeled due to missing radiation, condensation or diurnal cycle parameters in the model. It is therefore likely, that some wind velocity and wind direction components might be under-represented in such cases. Those local wind systems on the other hand usually make a rather small contribution to the overall wind quantity, which is normally dominated by adiabatic wind flow and fall winds. Adiabatic wind flow and fall winds in turn are computed correctly and in full by the GWS® MESO model. In any case, GWS® MESO wind maps do not show a tendency towards lower or higher wind values in comparison to real wind measurements.