Die  Kuste,  81  (2014),  273-290

280

Cuxhaven  2008  -  high  water

Cuxhaven  2008  -  low  water

30

^25
>.
£  20
OJ
I 15

<-80  -60  -40  -20  0  20  40  60  >80
difference  [cm]
(BSH-HBM  -  measurement)

30

^25
>.
g  20
0)
I 15

_ra
<u  5

<-80  -60  -40  -20  0  20  40  60  >80
difference  [cm]
(BSH-HBM  -  measurement)

Figure  5:  Frequency  distribution  of  high  (left)  and  low  (right)  water  level  differences  at  station
Cuxhaven  for  the  year  2008.

3.1.2  Baltic  Sea
In  the  Baltic  Sea,  water  level  variations  are  caused  mainly  by  wind  effects  and  seiches,
whereas  die  tidal  signal  is  relatively  small,  so  that  water  level  predictions  for  the  Baltic
differ  markedly  from  those  for  die  Nordi  Sea.  Therefore,  die  direct  model-water  level
output  is  validated  here.  As  it  can  be  seen  from  the  examples  presented  in  Fig.  6,  in  which
the  results  at  station  Warnemuende  are  shown,  die  model  describes  bodi  die  absolute
water  level  elevation  and  the  variability  of  it  rather  good.

11/03/08  15/03/08  18/03/08  23/03/08  27/03/08  31/03/08  04/04/08
Date

Figure  6:  Water  level  time  series  during  two  storm  events  in  2008  at  Warnemuende  station.  Observations ­
  are  shown  in  black,  BSH-HBM  data  in  blue.
Tab.  3  indicates  that  the  results  at  Warnemuende  are  representative  for  all  German  stations ­
  where  correlations  of  about  90  %  and  RMSDs  of  about  10  cm  are  to  be  found.
Moreover  the  simulated  standard  deviation  equals  nearly  the  observed  one  in  all  places.