DSD-INT 2020 DANUBIUS-RI Modelling Node Webinar

DANUBIUS-RI Modelling Node Webinar
2020 International Delft Software Days
December 1st 2020
Welcome!
Debora Bellafiore (ISMAR-CNR), Georg Umgiesser (ISMAR-CNR),
Francesca De Pascalis (ISMAR-CNR), Joost Icke (Deltares) and Björn
Backeberg (Deltares)

2020 International Delft Software Days - DANUBIUS-RI Modelling Node Webinar December 1st 2020
Code of conduct
•
• When listening…
• When talking…
• To ask a question…
• To participate in panel discussion
• Recordings…
2
Recordings will be available on https://softwaredays.deltares.nl/2020/index.php/agenda/webinars
…presenting…
Undock pane from Control Panel

DANUBIUS-RI The International Center for Advanced Studies on River –Delta –
Sea Systems is a pan-European distributed research infrastructure supporting
interdisciplinary research on large river-sea systems
RIVER-
ESTUARIES
RIVER-DELTAS
CULTURAL
HERITAGE
INDUSTRIAL
INFRASTRUTURE
ENERGY
GEO-HAZARDS
HABITAT
PRESSURES
NAVIGATION
DAMS
POLLUTION
RECREATION
FISHING
FLOODING
DROUGHT
CATCHMENT
CHANGE
Challenges to be addressed
▪ ensuring sufficient flows of water and sediment;
▪ maintaining structural integrity and continuity;
▪ preserve biodiversity;
▪ facilitating processes of natural self-organisation
over time;
▪ promoting resilience to extreme events.
RESEARCH AREAS
Global Change and Megatrends
Water and Sediment
Biodiversity and Ecosystems
Multiple Impacts on River- Sea Systems
Our Mission
• to provide state-of-the art research infrastructure from river
source to sea;
• to facilitate excellent interdisciplinary science;
• to offer integrated knowledge to manage and protect River-Sea
Systems.
Our Goals
• to overcome the current fragmentation of science, knowledge,
data & management in rivers and seas by integrating spatial,
temporal, disciplinary and sectorial thinking;
• to provide scientific solutions to environmental and societal
risks from climate change;
• to resolve problems arising from human impacts on River-Sea
Systems by using an interdisciplinary perspective, from source to
sea.
Our Vision
to achieve healthy River-Sea Systems and to advance their sustainable use, in order to live within the
planet’s ecological limits by 2050.

data information knowledge impact
Supersites
(living labs)
All over Europe
Hub
Data
Centre
Technology
Transfer
Officedata
DANUBIUS-RI The International Center for Advanced Studies on River –Delta –
Sea Systems is a pan-European distributed research infrastructure supporting
interdisciplinary research on large river-sea systems

The DANUBIUS-RI Modelling Node
Pool of
experts for
software
development
Pool of experts
for service
development
Working groups established to build on
community needs and develop the Modeling
Node potential
WG1 Spatially integrated models in RDS
systems
WG2 Interchange formats, standardization of
input/output
WG3 Post processing
WG4 Water quality and pollution modeling

Wish list for outcome
- Increase the Modelling Node Community, new involvement in WGs
- User needs identified and refined list of priorities
- Sharing of best practices and discussion on Modelling «PROBLEM SOLVING»
- Hypotheses on next opportunities for the modeling community in addressing RSS (projects, test cases,
collaborations)
- Identification of “hot topics” in the modelling community

December 1st 2020
Lightning Talks

December 1st 2020
Coupling numerical models
of two Supersites:
Middle Rhine and Rhine-Meuse Delta
Thomas Brudy

Supersite Middle Rhine:
▪ Free-flowing stretch of River Rhine from
Iffezheim to the Dutch border (~ 500 km)
▪ Hosted by BAW
Two supersites /
One River-Sea-System
▪ Common challenges, scientific issues:
- Sediment budget, management
- Sustainable navigation and
waterways
- Climate change
- …
Supersite Rhine-Meuse Delta:
▪ Delta region downstream German border /
medium-sized river Meuse / North Sea
▪ Hosted by RWS
https://de.wikipedia.org/wiki/Rhein-Maas-Delta

Coupling of the numerical models of two supersites
▪ Matching the strategic objectives of DANUBIUS-RI:
- Integrated understanding on the functioning of the River-Sea Systems
- Transboundary, interdisciplinary research on the River-Sea continuum
▪ Objectives in detail:
- Medium-term: external coupling of the hydrodynamic models
- Long-term: continuous transboundary models, morphodynamic models

Progress and results:
▪ Dutch/German-working group:
▪ Established by the shipping authorities
▪ Focus on numerical models in terms of
sediment management and navigation
▪ Completing the chain of hydrodynamic 2D-models
based on FE code TELEMAC
▪ Development of a new high performance code
for shallow water flow based on “Dumux”

Challenges and bottlenecks:
▪ Huge modelling areas => High Performance Computing
▪ Interfaces between the different model types
▪ Harmonization of modelling data:
- Coordinate reference system
- Height reference systems
https://www.killetsoft.de

Proposed future work:
▪ Morphodynamic 1D-model covering Lower Rhine / Waal
▪ Extension of the hydrodynamic 2D-model to the Netherlands
Panerdensch Kanaal / Waal
▪ Further development new code “Dumux”:
- Performance (fully implicit)
- Morphodynamic model
▪ Close collaboration between GER-NL / supersite coordinators
Sediment nourishment at Lower Rhine

December 1st 2020
Challenges in cross-scale modelling of
Elbe Estuary and German Bight
Joanna Staneva, Johannes Pein, Benjamin Jacob, Emil Stanev , Corinna Schrum (HZG)
joanna.staneva@hzg.de

Introduction and background of activity being presented
• Novel approaches are currently needed to understanding the
general context of climate change and the coastal hazards
• Increased interest in reducing prediction errors at coastal
scales.
• Extreme weather events in the marine realm.
• Bridging novel EO data and high resolution coastal models.
• Impact on climate studies and coastal predictions
• Wider social impact
Ems
Weser Elbe
Is coastal ocean modelling nowadays mature enough to address practical issues
from coastal engineering, search and rescue, BGC, green energy, coastal
management, response to climate change?

Objective and novelty of activity
• Implemented a coupled system for the three estuaries of te German Bight using unstructured grids to
address the efficient downscaling of coastal processes and the incorporation of freshwater land
discharges.
• A generic methodology is p for consistent interfacing (nesting, downscaling and transformed variables)
between large scale, structured and coastal models.
• Linkage with hydrology - experiments
• Prepared a set of criteria to extend CMEMS products towards the active coastal fringe, addressing the
requirements for boundary conditions, driving terms and derived fields.
• Interfaces towards operationally
• This quest for improved predictions is a repeated requirement by coastal stakeholders, in dire need of
more reliable data for currents, waves, surges, and water quality.

Progress and results:
Water quality example
Organic sediment Near-bottom dissolved oxygen conc.
• Numerical model delivers a realistic picture of the
dominant estuarine processes, e.g. eutrophication
• Model state variables can be used to derive
downstream products, e.g. risk maps for water quality
parameters like oxygen concentration (risk of hypoxia)

Challenges and bottlenecks faced
??
?? ??
- Synergy between newly available data and coastal models
- Solve practical problems
- Operational design
- Demonstration/ End-users/Society demands
Diversity of model components
Interaction between different models is important
Coastal and estuaries models (incl. BGC) are of not much interest if not well linked to
a) Larger-scale models
b) Atmosphere-wave models
c) Hydrology
Model diversity (Different design/coding)
Wave Model
BGC: Near-Coastal
models/applicationsCirculation model

Proposed future work:
Elbe esruary water quality : GCOAST (SCHISM- WWM-ECOSMO-SED)
Further integrated with sediment dynamics and morphodynamics
Pollution modelling – land-ocean continuum - atmosphere and ocean
Synergy of newly available observations and Ugrid estuary model – OSSE
Climate scenario (natural variability and human induced changes)

December 1st 2020
Coupled hydraulic/fish habitat
modeling in lower Nestos river
G. SYLAIOS, I. BOSKIDIS, N. KOKKOS, A. SAPOUNIDIS & E. KOUTRAKIS

Presently, the management of Heavily Modified Water Bodies (HMWB), like the Dammed
River Systems requires the assessment of the minimum Environmental Flow Requirement
(EFR) and its redefinition under conditions of climate change and water deficiency.
The EFR is defined as the minimum quantity, time availability and quality of river runoff
required to ensure the protection of internal and transitional ecosystems, as well as the
sustainable presence and development of humans, whose activities depend on these
ecosystems (Brisbane Declaration, 2007).

Aim of the present work is the implementation of a coupled hydraulic and fish habitat modeling of
Nestos River (Northern Greece) under various low flow conditions, serving for the analysis and the
ultimate determination of environmental flow at this sensitive river zone.
RIVER2D is a two-dimensional, depth averaged hydrodynamic
and fish habitat model, widely used in environmental flow
assessment studies.
Fish habitat modelling was conducted for four different fish
species (Alburnoides strymonicus, Barbus strumicae, Rhodeus
amarus and Squalius orpheus), divided into two age groups
(adults and juveniles), forced under five typical low flow
conditions.
Novelty: The work illustrates that ecohydraulic modelling of
river basins should be considered as an indispensable
component in sustainable water resources management, in
line to WFD implementation.

RIVER2D model uses the Weighted Usable Area (WUA) methodology calculated as the sum of the
Composite Suitability Index (CSI with a range of 0.0–1.0) estimated at each point of the computational
grid.
All examined fish species presented an increase of their WUA as the discharge increases. Juveniles are
more sensitive to low river flow conditions. Overall, the ecohydraulic model results illustrate that higher
river flows improve the quality and quantity of fish habitat in the river, while lower flows may stress it.

- Need for better DEM
- Better knowledge of Suitability Indices per parameter
- Upstream boundary conditions should be better defined

Conclusions
-In lower Nestos River the increase of base flow from 7 m3/s to 10 m3/s increased the habitat area
of all fish species (juveniles by 8.8% and adults by 15.3%).
-Squalius orpheus is the most sensitive species among adult age groups in terms of habitat area
response to low river flow conditions.
- The application of coupled hydraulic/fish habitat models in HMWB could be a valuable tool in
evaluating the impacts of low flow regimes on fish populations and assessing the minimum required
environmental flow in rivers.
Proposed future work
- Installation of meteohydraulic station equipped with an underwater camera with motion sensor to
assess fish behavior under variable river flow conditions.

December 1st 2020
Overview of Delft3D modeling of sediment
dynamics and morphodynamics of the
Scheldt estuary
Jebbe van der Werf, Bjorn Röbke, Thijs van Kessel, Mick van der Wegen,
Marcel Taal and many others

Scheldt estuary
Mouth ~25 km
+
Western Scheldt ~60 km
+
Lower Sea Scheldt ~40 km
+
Upper Sea Scheldt ~60 km
Well-mixed with average river
discharge ~100 m3/s
Meso- to macrotidal with tidal
range >5 m at Antwerp

Numerical models can contribute to policy & management
Potentially conflicting functions: safety, nature & navigation
Entrance to Port of Antwerp (BE) over NL territory
No free entrance between 1587 and 1795!

Modeling challenges
1. Complex physics: sand & mud, tides, winds & waves, transition from fresh water to sea water
2. Large domain (100+ km), long time scales (up to decades)
3. High spatial resolution required at beaches and intertidal areas (~10 m)
Percentage mud in the bed

Traditional approach: suite of fit-for-purpose models
Van der Werf et al., 2020, https://doi.org/10.1016/j.csr.2019.104027
Depth-averaged (2DH) Delft3D model for large-scale hydro-morphodynamics

Traditional approach: suite of fit-for-purpose models
Detailed 3D model for short-term sediment dynamics near Port of Zeebrugge
Van Maren et al., 2020, https://doi.org/10.1016/j.margeo.2020.106186

Future modeling: addressing the challenges
1. Flexible Mesh: easy local refining & faster computation
2. Fully coupled, long-term & large-scale sand-mud-morphodynamics
Intertidal area Western Scheldt: in 3 steps from rectangular grid
with resolution ~100 m to triangular grid with resolution ~15 m

DSD-INT 2020 DANUBIUS-RI Modelling Node Webinar

December 1st 2020
Towards a unified UGRID-NetCDF file format
for flow simulations on unstructured grids
Aissa Sehili, Günther Lang and Julia Benndorf

Background
UGRIDCFNetCDFfileformat
Preprocessing
Mesh – Initial conditions - Forcing
Simulation
UnTRIM Delft3D FM DELWAQ …
Results (synoptic)
Postprocessing
Aggregation – Analysis - Differences
Visualization
Unstructured grids are widely used in ocean,
coastal and estuarine modelling
Data exchange between models is still being a
major bottleneck requiring in some cases
considerable effort
Interpolation of imported data between edges,
nodes and faces is common resulting in non-
negligible accuracy loss

Background
UGRIDCFNetCDFfileformat
Preprocessing
Mesh – Initial conditions - Forcing
Simulation
UnTRIM Delft3D FM DELWAQ …
Results (synoptic)
Postprocessing
Aggregation – Analysis - Differences
Visualization
Need for a file format based on cross-platform
standards that is safe, robust and flexible
allowing data handling at Postprocessing level
NetCDF is a well established and freely
available cross-platform API
The CF (Climate and Forecast) conventions
metadata provide standards for the description
of NetCDF data and is widely used in the ocean
and atmospheric modelling community
UGRID can be regarded as an extension to CF
ready made for use with unstructured grids like
UnTRIM grids

Objective and goal of activity
Why UGRID CF NetCDF?
Safe use of dimensional data thanks CF canonical units
→ attribute “units”
Safe data transformation (e.g. intensive to extensive variables)
→ attribute “cell_methods”
Filtering based on number of data (easy tidal analysis)
→ attribute “ancillary_variables” + attribute “standard_name = number_of_observations”
Keep track of data processing activities
→ attribute “history”
Visualization of data at their original locations

Progress and results
UGRID-NETCDF
Output (HIS or MAP)
DCSM FM
(Deltares)
Forcing
GBOM (BAW)
Operational
Forecast
UGRID-NETCDF
2D/3D (loc,sec,area)
UGRID-NETCDF
BAW-Postprocessing
UGRID-NETCDF
BAW-Postprocessing
UGRID-NETCDF GRID UGRID-NETCDF GRID
The German Bight Operational Model.
Operational data communication between DFlow FM (Deltares) and UnTRIM (BAW)

float tnw_mean (TimePeriod, nFace) ;
:cell_methods = "TimePeriod: mean nFace: mean" ;
:ancillary_variables = “tnw_number of events" ;
int tnw_number of events(TimePeriod,nFace) ;
:units = “1" ;
:standard_name = number_of_observations" ;
Remark: “number of observations” (NOO) can be used to modify
visualization of results. E. g. results are shown
− only for locations where all events were detected, or
− for locations where at a certain NOO. were detected,
− … and so on.
Mean tidal low water at Elbe Estuary for the period 16.02-03.03.2019 (28 tides)
Tidal Characteristic Numbers of Water Level
https://wiki.baw.de/en/index.php/NCANALYSE
bathymetry [mNHN]
mean LW [m]

Main challenges
Implementation of UGRID methods in low level libraries in a generic way
Problems and compatibility issues from pre-requisite libraries (e.g. HDF5, PETSC,…)
Ongoing
Joint project between Deltares and BAW to enhance DFlow FM outputs to the
UGRID CF NetCDF standard used at BAW
DFlow FM NetCDF results should be processed without modification by BAW
NC-Postprocessors
Striving to
Make UGRID extensions full part of the CF NetCDF conventions
Backward compatibility: Use of UnTRIM results as is in DFlow FM
Simulations/Postprocessing

Open for further collaboration with the aim of
establishing UGRID CF NetCDF as a widely used
standard for data exchange and processing of
unstructured model results
UGRID
NetCDF
UnTRIM
Delft3D
FM
Others
Others

December 1st 2020
Post-processing of modelling data: tidal,
harmonic and long-term analysis and
aggregation
Julia Benndorf, Günther Lang, Aissa Sehili

Introduction and Objective
“For a better understanding of natural processes just
watching them is normally by far not sufficient. It may
be a much better approach to take, in a first step, some
precise and objective measurements. Afterwards, with
some characteristics of the processes in mind, different
analyses may yield characteristic numbers for each of
them.”
(from https://wiki.baw.de/en/index.php/Elementary_Aspects_of_Data_Analyses )
Reliable methods necessary in order to
• Analyse and aggregate large amounts of data from
modelling (e.g. event-driven for tides)
• Compare data from different models or systems
• Describe changes in time and space

Meantidalrange[m]
Parameter Tidal
Analysis
Long-term
Analysis
Harmonic
Analysis
Tidal Transport
Analysis
Long-term
Transport Analysis
Water level
Current velocity
Salt
Temperature
Suspended Sediment
Tracer
Oxygen
Bedload
Effective bed shear
stress
Morphodynamics
Energy
Progress and Results
https://wiki.baw.de/en/index.php/Analysis_of_Calculated_Results
Flood:Ebbcurrentduration
UnTrim (chunking) analysis
Duration of high SSC [h]Duration of high SSC [d]

Progress and Results
94 m³/s
69 m³/s
8 m³/s
Time of aggregation = 2010/04/26 22:55 to 2010/04/29 00:35
• Spacial aggregation for the
analysis of mean/max/pecentiles
within specific regions and
transport between them
• Here: water-volume flux into and
out of the Dollart area
(Note that there is a water level difference between
begin and end time of aggregation, therefore the net
water flux is not zero!)
Figure by Jens Jürges (Calibration of Ems-model), 2014. Modified by G. Lang, 2020.
water-volume flux
over edges between
control volumes
UnTrim (chunking) aggregation analysis

„During automatic calculation of various characteristic numbers several parameters must be properly chosen.
Inadequate choice of parameters may result in either wrong results or failure of data analyses programs […]. In
any case automatically calculated results shall be checked with respect to plausibility and consistence with
underlying physical principles. “
(from https://wiki.baw.de/en/index.php/Parameters_for_Data_Analyses )
• Choice of adequate analysis time and reference points
• Communication of results
• Different post-processing methods lead to different results
• Quality control

• Developments of standards for post-processing
• Validation and quality control of post-processing methods
• Open for discussion
• Participants in workpackage wanted!
Num. Model
Analysis
Aggregation
Visualization
Communication
of the results

December 1st 2020
Modelling mercury dynamic in lagoons
Coupling SHYFEM with MERC
Donata Melaku Canu, Ginevra Rosati
National Institute of Oceanography and Applied Geophysics -OGS

Introduction and background of activity being presented
•Mercury hot-spots in the Mediterranean and Black Sea
•Bioaccumulation and toxicity
•Cost of monitoring and measure
•Which is the relative contribute of the different processes
characterising the Hg cycle at the hot-spot level? management
support

Objective and novelty of activity
Novelty:
• Linking biogeochemical and transport processes in
transitional systems
• Linking abiotic and biotic components
• Providing information on the potential harm for
biota and human health (MeHg bioaccumulation
through seafood consumption)
Objective:
• Mercury cycle: biogeochemical model structure and implementation;
• Model results: HgII Hg0 and MeHg concentrations in water and sediment,
and budgets
• Model scenarios

Coupling the Shyfem Model with a Hg module and a
simple sediment transport module
* Data: Mela project (2005)
Water HgII concentrations

Silt in the Sea Bed HgII in the Sea Bed MeHg in the Sea Bed
Coupling the Shyfem Model with a Hg module and a
simple sediment transport module
December 2005 simulation

• Implementation, coupling and testing of two new
modules (sediment module)
• Limited data availability for Hg species (e.g. MeHg
concentrations)
• Uncertainities in the Hg cycle (e.g. transformation
rates)
• High computational cost

• Dynamic coupling with biological processes (e.g. bacterial POM remineralization)
• Adding bioaccumulation process for plankton (high variability across ecosystems, major driver to fish
concentrations)
• Dynamic coupling with the atmosphere
• Sediment diagenesis to improve benthic flux (?)

Jens Wyrwa
German Federal Institute of Hydrology , Koblenz
Department of Microbiol Ecology
Wyrwa@bafg.de
Modelling Water Quality in German Waterways
International Delft Software Days
December 1st 2020

since 1979 oxygen deficits main question
application to nearly all navigable rivers in Germany
deterministic model, main components:
❖primary production of organic carbon (algae)
❖mineralisation of organic carbon by bacteria
• second trophic level, algae consumtion
• balances of potentially limiting nutrients (N, P, Si)
• temperature, light, surface aeration
https://www.bafg.de/DE/08_Ref/U2/01_mikrobiologie/QSIM/qsim_node.html
german waterways

Elbe Estuary , July 7th 2011 12:03, ebb flow
water age (d)
Hamburg
river
inflow
north
sea
QSim used for:
large german rivers
specific:
• climat conditions
• fresh water
• retentions time
2 days … 2 months
Schöl, A., Hein, B., Wyrwa, J.& V. Kirchesch (2014):
Modelling Water Quality in the Elbe and its Estuary –
Large Scale and Long Term Applications with Focus on
the Oxygen Budget of the Estuary. Die Küste, 81
(2014), 203-232.

UnTRIM² offline or soft-coupling is faster if:
* multiple quality simulations
based on one flow field
(one year per night)
* interdisciplinary cooperation
Graphical User Interface GUI needed for:
• efficient model use
• safe model build up
• reproducible research
• quick viewing of results
GUI
flow
simulation
transport
metabolism
viewing
results
QSim software family
offline
coupling

Challenges
➢ Water quality simulation needs large model areas and integrated models
➢ Water quality modelling needs interdisciplinary expertise
➢ format standardisation would be a great help (WG2)
➢ tools for efficient and safe application are more essential in water quality modelling
(extend WG3?)
➢ biological simultion algorithms are more site specific than physical ones (WG4)

December 1st 2020
Break
We will be back for the open discussion at 11h45 CET

December 1st 2020
Open discussion

December 1st 2020
Wrap-up

December 1st 2020
Contact details
Debora Bellafiore debora.bellafiore@ve.ismar.cnr.it
Georg Umgiesser georg.umgiesser@ismar.cnr.it
Francesca De Pascalis francesca.depascalis@ve.ismar.cnr.it
Joost Icke Joost.Icke@deltares.nl
Björn Backeberg Bjorn.Backeberg@deltares.nl

DSD-INT 2020 DANUBIUS-RI Modelling Node Webinar

More Related Content

What's hot (20)

Similar to DSD-INT 2020 DANUBIUS-RI Modelling Node Webinar (20)

More from Deltares (20)

Recently uploaded (20)

DSD-INT 2020 DANUBIUS-RI Modelling Node Webinar