Low (web) Quality - BALTEX
Low (web) Quality - BALTEX
Low (web) Quality - BALTEX
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
83<br />
Does dynamical downscaling with regional atmospheric models add value to<br />
surface marine wind speed from re-analyses?<br />
Jörg Winterfeldt and Ralf Weisse<br />
Institute of Coastal Research, GKSS Research Centre, Geesthacht, Germany, ralf.weisse@gkss.de<br />
1. Assessment of added value using buoy wind<br />
The hindcast surface marine wind speed fields from the<br />
regional atmospheric model REMO in two configurations<br />
and the regional atmospheric model CLM are investigated<br />
with regard to their added value in comparison to the driving<br />
wind field from the NCEP/NCAR reanalysis (NRA_R1). To<br />
do so in a first step wind speed measurements from buoys,<br />
light ships and platforms in the eastern North Atlantic are<br />
considered as “truth”. Added value from the regional models<br />
is obtained when correspondence with both the measured<br />
statistical distribution and instantaneous wind speeds is<br />
higher than that of the reanalysis. Wind speed fields from<br />
NRA_R1, REMO and CLM are bilinearily interpolated to<br />
measurement locations and statistically compared.<br />
Winterfeldt and Weisse (2009) demonstrated that for<br />
instantaneous wind speeds the regional models do not have<br />
an added value both in "open ocean" areas and the German<br />
Bight. However, in the English Channel, where local<br />
topography and associated local wind regimes become<br />
important, the regional model shows an added value for<br />
instantaneous wind speeds.<br />
Concerning the wind speed distribution there's a clear<br />
indication for an added value of the regional models in<br />
coastal regions, especially for higher wind speed percentiles,<br />
while in "open ocean" areas the NRA_R1 is better reflecting<br />
observed distributions.<br />
2. Assessment of added value using QuikSCAT<br />
wind speed retrievals<br />
The added value of the dynamically downscaled wind was<br />
assessed with satellite data, namely QuikSCAT Level 2B<br />
12.5 km (L2B12) wind speed retrievals following the<br />
thorough validation of the quality of L2B12 data with buoy<br />
winds in the eastern North Atlantic (RMSE: 1.5 m/s) by<br />
Winterfeldt et al. (2009). For that purpose L2B12, REMO<br />
and global NCEP/NCAR reanalysis (NRA_R1) data were<br />
co-located for the years 1999-2007. Co-location criteria<br />
between REMO and L2B12 data are within 0.1° and 0.06° in<br />
longitudinal and latitudinal distance from REMO model grid<br />
points and within 20 minutes, leading to a high quality<br />
L2B12 gridded wind speed data set with REMO grid<br />
dimensions. NRA_R1 data was interpolated in time and<br />
space onto the REMO grid.<br />
Figure 1. Modified Brier Skill Score calculated from colocations<br />
between QuikSCAT L2B12, NRA_R1 and SN-<br />
REMO (SN stands for use of spectral nudging) in the wind<br />
speed range from 3 to 20 ms-1 and the years 2000 to 2007,<br />
where QuikSCAT L2B12 serves as ``truth``, NRA_R1 as<br />
reference ``forecast`` and SN-REMO as ``forecast``. Blue<br />
areas indicate value lost, while red areas indicate value<br />
added by dynamical downscaling.<br />
3. Seasonal variability of added value<br />
Strong interseasonal differences exist, in winter enhanced<br />
cyclonic and meso-cyclonic activity increases the potential<br />
of dynamical downscaling. In winter time the added value<br />
is more pronounced around Iceland and Greenland, south<br />
of Iceland and within the Gulf of Lyon/Mistral region (see<br />
Fig. 2 and 3).<br />
Summarizing QuikSCAT is a valuable tool to identify<br />
marine regions where dynamical downscaling of wind<br />
speed makes sense.<br />
A slightly modified Brier Skill Score (BSS) was used to test<br />
to what extent the regionally modelled wind gives a better<br />
reproduction of QuikSCAT wind speed than the NRA_R1.<br />
The modified version of the BSS varies between -1 and +1<br />
and simplifies the comparability of positive (value added)<br />
and negative (value lost) scores.<br />
Fig. 1 confirms the point stated by Winterfeldt and Weisse<br />
(2009) for a wide area including the eastern North Atlantic,<br />
the Baltic, Mediterranean and Black Seas: dynamical<br />
downscaling does not add value to NRA_R1 wind speed in<br />
open ocean areas (blue), while it does for complex coastal<br />
areas (red).<br />
Figure 2: As Figure 1 but for Summer (JJA).