Presentation by Emil Stanev (HZG Institute of Coastal Research, Germany), at the DANUBIUS Modelling Workshop, during Delft Software Days - Edition 2019. Friday, 8 November 2019, Delft.
DSD-INT 2019 Elbe Estuary Modelling Case Studies-Stanev
1. Elbe Estuary. Modelling Case Studies
Emil V. Stanev
DANUBIUS Modelling Workshop
Friday, 08 November 2019,
Delft, The Netherlands
(Elbe Delta)
2. Motivation
1. Share knowledge and experience about the modelling of river-sea systems.
2. Consider possible research-support when solving society-relevant issues,
climate change and environmental protection
Focus on
1. What happens when the river water encounters the sea water?
2. Nesting and seamless approach; coupling different modules
Outline
1. Introduction
2. SCHISM
3. SED3D
4. WWM
5. ECOSMO
3. 1. Introduction (challenges)
The water cycle
Global European Seas
J.P. Peixoto & A.H. Oort, Physics of Climate, 1992
Stanev and Lu (2013)
4. Chl-a
µg/l
Remote Sensing & FerryBox Data
1°E 2°E 3°E 4°E 5°E 6°E 7°E 8°E
54°N
52°N
FerryBox
Sediment dynamics and biogeochemical processes
W. Schröder, personal communication
5. W. Schröder, personal communication
Vertical mixing: a
fundamental
ingredient of
coastal models
6. Where the river meets the sea
Different estuaries and mixing processes
• How is the "salt balance" maintained in
an estuary?
• Why does the river water not simply
flush out all the salt and turn the
estuary into fresh water?
Strongly, partially mixed, salt wedge estuaries
7. Hansen & Rattray, (J. Marine Research, 23, 104-122; 1965)
CD is the friction parameter,
UT is the amplitude of the depth-averaged tidal velocity,
ω is the tidal frequency,
No = (βgsocean/H)1/2 is the buoyancy frequency,
H is the depth,
UR is the velocity of river flow,
ß= 7.7 × 10-4 is the coefficient of salinity contraction,
and g is the gravitational acceleration.
M2 = CDUT
2/(ωNo H2)
Estuarine classification
Geyer and MacCready
(Annu. Rev. Fluid Mech. 2014. 46:175–97)
M2 = CDUT
2/(ωNo H2) quantifies the effectiveness of tidal
mixing measuring the ratio of the tidal timescale to the
vertical mixing timescale.
Frf = UR/(ßgsoceanH)1/2, freshwater
Froude number measuring the
ratio between the net velocity due
to river flow and the maximum
frontal propagation speed No H.
8. Hydrology
HD-Modell
2. The model
GCOAST Modelling system
Waves
WAM
Atmosphere
COSMO-CCLM
Ocean
NEMO/SCHISM
Biogeochemistry
ECOSMO/E2E
Atm. Chemistry
CMAQ
Marine
chemistry
MECOSMO
Bio-
accumulation
Drift
Models
Coupler
OASIS
SPM
9. 3D, primitive equations, unstructured-grid.
- Upgrade from an existing model (SELFE, A Semi-implicit
Eulerian-Lagrangian Finite Element model for cross-scale
ocean circulation).
- Uses hybrid finite element and finite volume approach.
- New viscosity formulation (effectively filters out spurious
modes without introducing excessive dissipation).
Semi-implicit Cross-scale Hydroscience
Integrated System
Model; www.schism.wiki
- New higher-order implicit advection
scheme for transport (TVD2) is proposed
to effectively handle a wide range of
Courant numbers
- Addition of quadrangular elements
into the model
- Flexible vertical grid system (Zhang et
al. 2015, OM)
- Model polymorphism that unifies
1D/2DH/2DV/3D cells in a single model
grid.
Zhang Y.J., F. Ye, E. V. Stanev, and S. Grashorn
(2016, Ocean Modelling).
13. 3. SPM dynamics
• A zone within which the
suspended sediment
concentrations are higher than
those in the river or further
down in the estuary.
• The turbidity maximum occurs
more often in well-mixed and
partially-mixed estuaries, and
less often in stratified estuaries.
• In many places the turbidity
maximum contains more
sediment than brought by rivers.
16. 4. Wind waves. WWM III and ist coupling with SCHISM
-WWM III (third generation spectral wave model) is described by
Roland et al. (2012).
-Wind input and dissipation is as in Bidlot et al. (2002).
-SCHISM gets from WWM III the radiation stress.
-The coupling between the wind-wave and the circulation model is
made through the friction velocity computed from the wave model
(Bertin et al., 2015).
-The source terms for depth-induced wave breaking and bottom
friction are computed as explained by Bertin et al. (2015).
Schloen et al. (2017, OMOD)
17. Model validation
Wind speed and significant wave height in July 2013
The meteorological situation
and wind waves
Observed and simulated significant wave height, direction and peak period
18. Left: Spatial and temporal variability of difference between
salinity simulated in experiment RWF and RF: Consequent
snapshots during one tidal period are shown for 23-24.07.
Velocity-difference vectors are also plotted. Right: Time versus
distance diagram along a section line north of the islands.
(a) The ratio of significant wave height and tidal range
averaged along the transect line north of the islands. (b)
Wind speed and direction along the section line. (c)
Difference of the u-component between experiment RWF
and RF.
Temporal-spatial variability
19. (a) The ratio of significant wave height and tidal range averaged for the whole model run and for a period
with high waves (23 Jul 2011). (b) Vertical profile of difference in salinity on transect along the tidal channel
for RWF and RF averaged over 16 tidal M2-periods.
Estuarine implications
20. Results
- Density gradients in the coastal zone reduce the tidal current by 18 %.
- Wind waves enhance the circulation in some cases. The latter happens when strong
winds blow resulting in long-shore currents following the western Dutch coast and
the German Wadden Sea islands.
- The wave-induced transport of salt leads to changes in the horizontal salinity
distribution, which are very pronounced in regions of fresh water influence.
- The weak stratification dominating the patterns of salinity in the coastal zone is
mostly destroyed by wind waves. Thus, effects created by wind waves tend to
substantially modify the estuarine circulation.
- More extended description of the results can be found in Schloen et al. (2017).
21. WGE
5. BGC modelling
Elbe model coupled hydrodynamically to bigger set-up (GB), atmospheric forcing
by DWD data, river discharge, nutrient and plankton load from observations
22. 22
Ecological module: ECOSMO2 – a
pelagic NPZD model plus bottom pool of
nutrients
Coupler: FABM (Bruggeman & Bolding,
2014) linking hydrodynamics and
ecology
ELBE BGC-MODELLING SYSTEM
Set-up characteristcs: 33k horizontal nodes, 20
s-layers, TVD² transport, KL-turbulence model.
ECOSMO/
MAECS/ SPM
FABM
Daewel & Schrum, 2013
Schrum et al., 2006
26. Conclusions
Coastal ocean modelling is nowadays mature enough to
address practical issues from coastal engineering,
search and rescue, BGC, green energy, coastal management,
response to climate change
Emil.Stanev@hzg.de