Fróðskaparrit - 01.01.2007, Page 131

SAMANBERING AV ALDUFORSØGNUM FRÁ ECMWF VIÐ ALDUMÁTINGAR
GJØRDAR Á FØROYALEIÐINI
129
however, tliat the initial state is never
known exactly. To overcome this problem
in operational forecasts, the initial state for
the forecast is obtained by a combination
of the latest forecast and all available new
observations. This procedure, where the
observations ‘push’ the initial state to-
wards reality (measurements at the given
time), is known as analysis.
The model results validated here are the
6-hourly analysis data from the ECMWF
operational archive of the limited area
wave model (wave parameters and wind
speed) and the atmospheric model (wind).
The integrated wave parameters generated
by this rnodel (e.g. Hm0, Tm(p etc.) are based
on the predicted wave spectrum, with a
padded ‘diagnostic’/'5 tail in the frequen-
cies above the model resolution. The wind
speed parameter from the wave model cor-
responds to the winds that are actually seen
by the wave model. In the case of the lim-
ited area wave model, it is essentially the
same wind as provided by the archived at-
mospheric model fíelds, except that the
data have been interpolated onto the wave
model grid. Therefore it does not contain
any values over land points (missing data).
When interpolating such fíeld, the closest
grid point is used whenever missing data
are present; otherwise bi-linear interpola-
tion is used.
Available metocean data
The validation is performed against all
suitable buoy data in the region. The non-
directional waverider buoys WV-1, WV-2,
and WV-3 are part of the Faroese buoy
measuring program which was started in
1979 (Davidsen and Hansen, 1981) and are
operated by the Office of Public Works
(OPW) on the Faroe Islands. The direc-
tional waverider buoy, WVD-4, is operated
by the company DataQuality, on the behalf
of a consortium of oil conrpanies (FOIB).
The meteorological buoy K7 (also known
as 64046) is operated by the MetOffice
UK. Data from the land based meteorolog-
ical station at F8 are also used. This station
is also operated by OPW. Table 1 gives the
locations of the measuring sites and the
closest respective rnodel points.
The periods of validation (Table 2) are
the four months that contain the highest
recorded wave heights at WVD-4 site so
far. The WVD-4 buoy is chosen as the
main reference as this site is less influ-
enced by sheltering and other fíne scale
features compared to the OPW buoys. The
WVD-4 buoy is also the only directional
buoy, and the only buoy where the com-
plete data-set (raw data etc.) are made
available. Another factor that also might
complicate the model comparison at some
of the OPW sites is the mooring type used
there (Niclasen and Simonsen, 2005b).
WV-l WV-2 WV-3 WVD-4 K7 F8
Site Lat. 61.80°N 61.87°N 62.50°N 61.30°N 60.50°N 61.97°N
Long. 6.22°W 7.53°W 6.83°W 6.28°W 4.90°W 6.75”W
Depth lOOm I30m 1 OOm 240m lOOOm —
Model Lat. 61.75°N 61.75°N 62.50°N 61.25°N 60.50°N 62.00°N
Long. 6.25°W 7.50°W 6.75°W 6.25°W 5.00°W 6.75°W
Table 1. Location and depth of the measuring sites and their respective closest model points.