808 
U. Callies et al.: Surface drifters in the inner German Bight 
Ocean Sci., 13, 799-827, 2017 
www.ocean-sci.net/13/799/2017/ 
(BSHcmod; Fig. Al), (b) Eulerian currents plus windage 
(BSHcmod + W; Fig. A2) or (c) Eulerian currents plus 
Stokes drift (BSHcmod+S; Fig. A3). Numerical data dis 
played in the graphs are provided in the Supplement. It ap 
pears that combining BSHcmod currents for a 5 m depth sur 
face layer with either windage or Stokes drift brings corre 
sponding simulations closer to both observations (Fig. 4) and 
simulations based on Eulerian surface currents from TRIM 
with 1 m vertical resolution (Fig. A4). A key effect of the in 
clusion of extra wind or wave effects is the intensification of 
westward transports in agreement with wind directions that 
occur most frequently. One should remember that achiev 
ing reasonable agreement between overall strengths of these 
transports in simulations and observations was the criterion 
which led to the specific values we assigned to a or fi in 
Eq. (1) (see Sect. 2.2.3). 
Analysing short-term drifter displacements on a daily ba 
sis enables a more detailed assessment of model perfor 
mance. Simulations of 25 h drift paths were initialized ev 
ery day on days 0-53 at 13:00UTC. Full sets of correspond 
ing plots are provided in the Supplement. The first collection 
(SMI in the Supplement) shows 25 h drifter displacements 
that were observed. The second (SM2 in the Supplement) 
compares corresponding simulations based on either BSHc 
mod or TRIM Eulerian currents with these observations. The 
third (SM3 in the Supplement) is similar except that BSHc 
mod currents are complemented by parametrized windage 
(BSHcmod + W). Finally, the fourth collection (SM4 in 
the Supplement) compares BSHcmod simulations includ 
ing either windage (BSHcmod + W) or Stokes drift (BSHc 
mod + S). 
Drawing on the material from SM3, Figs. 7-9 present re 
sults for 12 selected days, comparing simulations based on 
TRIM surface currents with those based on BSHcmod + W. 
Each panel combines all drifters that are available at the re 
spective time. Observed drifter displacements are coloured in 
agreement with Table S1. 
Concentrating on the four drifters that travelled longest, 
bars in Fig. 10a show daily values of separation between ob 
served and simulated end points of 25 h drift paths, referring 
again to simulations with either TRIM or BSHcmod + W. To 
show the relative importance of drift errors, total distances 
covered according to observations or simulations are also in 
cluded. Figure lOd shows the angles between observed and 
simulated drifter displacements. Time series (25 h means) of 
wind speeds used in TRIM and BSHcmod (and also BSHc 
mod + W) are shown in Fig. 10b together with surface Stokes 
drifts from wave model WAM. Figure 10c copies observed 
Heligoland wind vectors from Fig. 3. 
The following description highlights some key aspects of 
drifter observations and concurrent simulations during differ 
ent sub-periods of the experiment. 
Days 0-6 (27 May-2 June): This is a period characterized 
by cyclonic residual currents increasing in strength 
(Fig. 3). Driven by winds mainly from south-west, 
drifters move fast towards a north-eastern sector. Sim 
ulated drift distances agree well with observations. Ap 
preciable errors for TRIM arise from moderate direc 
tional deviations in combination with large displace 
ments (Fig. If). 
On day 2, neither model simulates the neighbouring 
drifter nos. 5 and 6 to move into different directions 
(SM3). On day 3 (Fig. 7a and e), only BSHcmod + W 
captures the deviant direction of drifter no. 7. Compar 
ing simulations based on BSHcmod + W (Fig. 7a) with 
those based on BSHcmod (SM2) reveals that the deviant 
simulation of drifter no. 7 arises from spatial variation 
of BSHcmod currents. By contrast, inclusion of more 
large-scale windage affects all drifters tracked in a very 
similar way. On day 6 (Fig. 7b and f), the again deviant 
movement of drifter no. 7 (now rotated to the opposite 
direction) is no longer reproduced by BSHcmod + W. 
For drifter no. 1, simulations are generally poor. On 
days 3-5, the drifter already enters the complex coastal 
bathymetry which is insufficiently resolved in both 
models (e.g. Fig. 7a and e). 
Days 7-10 (3-6 June): The strong cyclonic regime de 
clines, strong south-west winds first cease and then 
blow from different directions (Fig. 3). Observed dis 
placements of drifter nos. 5, 6 and 8 take a minimum 
on days 7 or 8 (Figs. 6 and 10a). On day 7, major di 
rectional errors occur under variable wind conditions 
(Figs. 7c, g and lOd). Only drifter no. 9 rotates its move 
ment from north-east to north-west already on day 7; all 
other drifters follow on day 8 (Fig. 7d and h). Speed 
of drifter no. 9 shows a strong peak on day 8 (Fig. 6). 
Observed drifter displacements seem to decrease with 
distance from the coast, a variation not resolved in sim 
ulations (Fig. 7d and h). Also considering Stokes drift 
does not help reproduce this spatial gradient (SM4). 
Sub-mesoscale differences in drift speed (e.g. day 8; 
Fig. 7d) and direction (e.g. day 10; Fig. 8a) giving rise 
to the fast convergence of drifter nos. 6 and 8 (Fig. 5) 
remain unresolved in both models. Neither model cap 
tures the special behaviour of drifter no. 7 which contin 
ues its fast movement towards northern directions (com 
pare Figs. A2d and A4d with Fig. 4d). 
Days 11-14 (7-10 June): Winds from the north-west or 
north trigger an anticyclonic circulation (Fig. 3). On 
day 11, the inclusion of windage greatly reduces er 
rors of BSHcmod simulations for drifter nos. 6, 8 and 
9 (compare Fig. 8b with SM2), mainly due to improved 
drift directions. Only for drifter no. 5, moving much 
slower despite its proximity to other drifters, adding 
windage leads to drift velocity on day 11 being greatly 
overestimated (Fig. 6).