Page 4 - ENERGIA_MARE

Author's personal copy

L. Liberti et al. / Renewable Energy 50 (2013) 938e949 941

Table 3
Statistics of satellite and model significant wave height (Hs) comparison for the
Adriatic Sea.

Satellite Samples Bias (m) Rmse (m) Slope si d

Topex-Poseidon 21,889 À0.256 0.407 0.763 0.438 0.891
Jason-1 9911 À0.108 0.436 0.897 0.440 0.895
Jason-2 9447 À0.107 0.404 0.906 0.417 0.896
Envisat À0.165 0.476 0.751 0.595 0.845
ERS-2 32,832 À0.289 0.475 0.753 0.485 0.815
24,043

shows the scatter plot of model and Jason-1 Hs over the entire Fig. 4. Locations of RON wave buoys. Filled symbols indicate buoys included in model
Mediterranean. Considering the number of sample points analyzed, validation. Dashed grid lines are 200 km apart.
and the fact that the lighter gray tiles represent bins containing less
than 5 samples, the data points dispersion shown in the plot is by qualitatively comparing the model and buoy time series. After
acceptable. Similar results are found for all the other satellites. As assessing the good quality of the results, more rigorous compari-
reported in [23] satellite measures of Hs tend to be less reliable for sons were carried out by constructing scatter plots of correspond-
values lower than 0.5 m a fact that might explain the higher ing model and buoy values and by computing the statistics given by
dispersion of the points in lower left section of the scatter plot. Equations (1)e(5). We excluded from the analysis buoy records
where the peak spectral period Tp fell in the infra-gravity waves
2.3. Validation against buoy data range, above 20 s.

Model results were also compared to buoy wave measurements Fig. 5 shows the Hs scatter plot for the Alghero buoy while
collected by the Italian Wave measuring Network (Rete Onda- Table 5 lists the values of the bias, si, slope, rmse and d for each
metrica Nazionale, RON) managed by Institute for Environmental buoy. By inspecting the values in Table 5 we observe that we can
Protection and Research (ISPRA). Eight directional deep water group the buoys in three subsets depending on the level of agree-
buoys are in operation since July 1989 recording significant wave ment between model and buoy values. The results obtained at
height Hs, peak period Tp, mean period Tm and mean wave direction Alghero, Crotone, Cetraro and Capo Gallo buoys are in very good
agreement with the data with best fit line slopes around the unity,
qm. Until 2002 wave parameters were normally recorded every 3 h. biases of the order of centimeters and scatter indices below 0.4. For
a second set of buoys including La Spezia, Ponza and Cetraro the
During storms, when Hs exceeded a buoy-specific threshold, the results appear to be still satisfactory but biases tend to be in the
recording rate was increased to 300, in 2002 the measurement rate order of 0.1 m and best fit line slopes are below 0.9. Hs appears to be
was set to 300 regardless of the sea state [26]. Wave measurements
are currently available for each buoy starting from its initial 8
deployment date up to mid 2008. In order to obtain a set of
statistically homogeneous samples we took into account only
three-hourly recorded data so that the measurements dating back
to 2001 and 2002 could be included in our analysis. This choice is
also justified by the 6-h temporal resolution of forcing wind fields
which is significantly lower than the 300 buoy measuring rate.
Buoys whose data record we judged too short were not used in the
analysis. Table 4 gives an overview of the RON buoys data avail-
ability while Fig. 4 shows the location of the RON buoys.

Buoy measures were compared to model output extracted from
the nearest computational node. A first validation was performed

Table 4 500
Data availability of RON wave buoys. Buoys indicated in bold were used for model
validation. 6 250

Buoy First record Last record 3 hr Expected Valid 100

records 3 hr records (%) 4 50

20

10
2

5

0
0

02468
Hs - Buoy
(m)

Fig. 5. Correlation between buoy and model Hs at Alghero. Dashed line is the best fit
line between model and buoy data points.
after records Hs - Model
(m)
1/1/2001
Entries
Alghero 01/07/1989 05/04/2008 15,283 21,210 72.1
16,827 80.2
Catania 01/07/1989 05/10/2006 12,549 19,093 78.4
18,240 60.0
Crotone 01/07/1989 15/07/2007 14,962 21,207 72.3
21,209 73.7
La Spezia 01/07/1989 31/03/2007 10,952 21,113 60.6
21,169 68.4
Mazara del Vallo 01/07/1989 04/04/2008 15,323 21,209 78.4
15,812 64.6
Monopoli 01/07/1989 05/04/2008 15,641 67.3
5664 72.5
Ortona 01/07/1989 24/03/2008 12,786 12,408 69.1
62.8
Ponza 01/07/1989 31/03/2008 14,479 7872 63.7
7872
Cetraro 28/02/1999 05/04/2008 16,630 3120

Ancona 10/03/1999 31/05/2006 10,212

Capo Comino 01/01/2004 12/09/2005 3813

Capo Gallo 01/01/2004 31/03/2008 9001

Capo Linaro 02/01/2004 12/09/2006 5441

Punta della Maestra 02/01/2004 24/11/2004 2616

Cagliari 06/02/2007 02/03/2008 1986