Page 5 - ENERGIA_MARE

Author's personal copy

942 L. Liberti et al. / Renewable Energy 50 (2013) 938e949

Table 5
Statistics of buoy and model significant wave height (Hs) comparison.

Buoy Bias (m) Rmse (m) Slope si d 0.6 Buoy
0.4 Model
Alghero À0.005 0.311 0.985 0.278 0.974 0.2
Ancona À0.214 0.361 0.725 0.477 0.889 Frequency
Catania À0.178 0.308 0.747 0.501 0.876
Crotone 0.276 0.993 0.374 0.949
La Spezia 0.004 0.283 0.851 0.354 0.951
Mazara del Vallo À0.143 0.257 1.022 0.253 0.971
Ortona 0.284 0.753 0.460 0.911
Ponza 0.013 0.273 0.892 0.328 0.953
Monopoli À0.150 0.307 0.836 0.427 0.917
Cetraro À0.103 0.241 0.897 0.341 0.960
Capo Gallo À0.124 0.255 1.040 0.339 0.962
À0.070

0.019

Table 6 0.0
Statistics of buoy and model wave average spectral period Tm comparison. 0 100 200 300
Average Wave Direction
Buoy Bias (s) Rmse (s) Slope si d (Deg.)

Alghero 0.230 0.791 1.032 0.172 0.915 Fig. 6. Frequency distribution of model and buoy average wave direction at Alghero.
Ancona À0.349 0.959 0.883 0.240 0.718 Only records with Hs > 1 m are considered. Incoming wave direction expressed in
Catania 1.163 0.996 0.274 0.725 degrees clockwise from the north.
Crotone 0.142 0.805 1.031 0.207 0.847
La Spezia 0.206 1.049 1.107 0.257 0.813 distribution since the wave power flux, as it will be shown in the
Mazara del Vallo 0.506 0.780 1.030 0.177 0.865 next section, is calculated using the significant wave period Te
Ortona 0.195 0.718 0.992 0.198 0.790 which is not readily available from the buoy measures. As shown in
Ponza 0.063 0.646 0.993 0.160 0.881 Table 6, the regression line slope is almost one for each buoy with
Monopoli 0.035 1.017 0.877 0.257 0.656 the exception of Monopoli and Ancona where the model under-
Cetraro À0.327 1.009 1.004 0.223 0.818 estimates Tm and the bias is negative. For all the remaining buoys
Capo Gallo 0.156 0.925 1.048 0.225 0.822 the model tends to overestimate the wave average period; where
0.307 the regression line slope exceeds the unity, this trend is more
marked at higher values of the period. The overestimation
generally underestimated for the buoys located in the Adriatic Sea. measured by the bias is relatively small. Bias values exceeding 0.5 s
This is probably due to the rather coarse resolution of the wind are found only at La Spezia. Rmse values are considerably higher
fields compared to the extent of the Adriatic Sea. As argued by [27], than the ones obtained for Hs and d values are lower implying that
the effect of the complex orography, which is not well described at model Tm values have higher dispersion. In any case, the absolute
the resolution of atmospheric model, plays in this case an impor- difference between buoy and model average period is less than 1 s
tant role in defining the wind fields. In fact, the comparison of in more than 70% of samples for all the buoys with the only
satellite altimeter data in the Adriatic Sea in Table 3 shows better exception of Catania.
agreement between model and measures when compared to the
buoys. Satellite tracks are located further offshore than the wave As a final validation step, following the approach used by Han-
buoys where the influence of the land is less important. A more son [28], we compared buoys and model mean wave directions.
accurate description of the wave energy potential in the Adriatic This analysis was carried out by calculating the directional bias and
Sea requires wind fields at higher temporal and spatial resolution, the directional variance of the difference between model and buoy
however, preliminary and conservative estimates of the wave
energy potential can still be obtained by the present model. Hs is Buoy
underestimated also at the Catania buoy. This buoy is located in 0.6 Model
a sheltered position characterized by a mild wave climate, two
factors that are known to contribute in reducing the model
performance.

We further validated the wave model by comparing buoy and
model average spectral period Tm. The average wave spectral period
Tm is used here as an estimate of wave spectra frequency

Table 7 Frequency 0.4
Circular statistics of buoy and model wave average spectral direction qm comparison.
0.2
Buoy Bias (Deg.) var
0.0
Alghero 4.51 0.036 0 100 200 300
Ancona 9.41 0.230 Average Wave Direction
Catania 14.81 0.053 (Deg.)
Crotone 8.09 0.085 Fig. 7. Same as Fig. 6 for Mazara del Vallo buoy.
La Spezia 2.94 0.056
Mazara del Vallo 11.00 0.057
Ortona 13.48 0.101
Ponza 8.76 0.116
Monopoli 5.00 0.121
Cetraro 6.39 0.063
Capo Gallo À5.21 0.026