Fróðskaparrit - 01.01.2007, Side 132

130
VALIDATION OF THE ECMWF ANALYSIS WAVE DATA
FOR THE AREA AROUND THE FAROE ISLANDS
Event number Start date End date
1 20/8-2000 20/9-2000
2 15/1-2002 15/2-2002
3 1/6-2002 1/7-2002
4 1/1-2003 1/2-2003
Table 2. Duration of the inspected events
The data from WVD-4 are quality
checked for all events by the data provider.
On top of tliis check, different nianual in-
spections of the raw data from the peak of
the storms are performed, which resulted
in rejection of a few measurements in the
January 2003 storm. Data from the other
locations are only quality checked by pa-
rameter behaviour; that is outliers are re-
moved if they are not supported by several
measurements etc.
The time tags connected to measured
wave parameter values have been adjusted
so they correspond to real-time, rather than
completion times of the measured time se-
ries (data-burst) which generate the wave
spectra.
Wave buoys can only accurately meas-
ure waves up to some fixed cut-off fre-
quency^. A, and the calculation procedures
of the buoys do not add a diagnostic upper
frequency tail above fhigh. For this reason
integrated wave parameters, especially
those based on higher order momentums of
the spectrum (e.g. Tm02), are affected by
this upper frequency limit of the buoy
measuring system. At WVD-4 fhj =0.58
Hz (Heinesen, 2005), at K.7 fhig=0.25 Hz
(Peter Fenna at Met Offíce UK, pers. com.,
2006) and at the OPW buoys fh h is esti-
mated to be in the region of 0.5 Hz (Heine-
sen, 2001).
In order to compensate for the effect of
fh h on the measured wave parameters, the
wave parameters from WVD-4 are recal-
culated from the measured spectrum
padded with an f5 high frequency tail. As
the spectra are not available for the OPW
buoys or the K7 buoy, it is not possible
compensate for the influence of the cut-off
frequency on the measured wave parame-
ters from these sites.
In relation to the wave measurements
from K7 it ought to be mentioned that the
wave height and wave period (labelled Hm0
and Tm02 here), are in fact calculated di-
rectly from the time series as root-mean-
square wave height and zero-crossing
wave period T_, and not from the measured
wave spectra as at the other buoys used
here. According to the operator, the differ-
ence between using the root-mean-square
value instead of the wave spectra to com-
pute the wave height has been found to be
insignificant (Peter Fenna at Met Offíce
UK, pers. com., 2006). The two wave peri-
ods Tm02 and T_ should in average be di-
rectly comparable as they both correspond
to the period of the mean wavelength, or
alternatively if the wave field is assumed
to be a random Gaussian process, then
these two periods have been shown to be
equal (Rice, 1944; 1945).
Wind measurements are available for
comparison from the meteorological sta-
tion F8 and the K7 buoy. Numerical inves-
tigations of stable circumstances suggest
that the measured wind speed at the F8 sta-
tion is on average comparable to the off-
shore 10-meter wind (Heinesen, 2002).
Nevertheless direct comparison of wind-