Presentation by Frédéric Soulignac (LEESU) at the Delft3D User Days, during Delft Software Days 2016. Thursday 3 November 2016, Delft.

1. 1: The French National Institute for Agricultural Research
Application of Delft3D to evaluate the spatio-
temporal variability of Lake Geneva
ecological status
Delft3D User Meeting - 03 November 2016
Frédéric Soulignac1, P.-A. Danis2,10, D. Bouffard3,
V. Chanudet4, E. Dambrine5, Y. Guenand6, B. Guillermin1,
T. Harmel7,10,2, B. Ibelings8, I. Kiefer3, D. Trevisan1,
R. Uittenbogaard9 and O. Anneville1
3 4 5 6 7 8 9 10
2: The French National Agency for Water and Aquatic
Environments

2. Lake Geneva
Image from CIPEL, 2016
Lake catchment (7 419 km2)
Catchment of the Rhône river
after Lake Geneva
Rivers
Border France-Switzerland
Lake Geneva
2
Frédéric Soulignac, Delft3D User Days
2016
Surface area = 580 km2
Max depth = 309 m
Residence time = 11 years
Introduction – Material and methods – Results and discussion – Conclusions
I work here!

3. • European Water Framework Directive (EU, 2000)
• 3 types of indicators: biological, physico-chemical and hydro-morphological
• In France (French technical guide for evaluating ecological status of lakes
and reservoirs, 2016 ; De Bortoli and Argillier, 2008)
– Annual evaluation based on 4 samples taken over the year and at 1 sampling
station
– Parameters
• Chlorophyll a concentration (Chla)
• Ammonium concentration (NH4)
• Nitrate concentration (NO3)
• Total phosphorus concentration (TP)
• Secchi depth (Secchi)
• And others
• Lake Geneva
Evaluation of Lake ecological status
Frédéric Soulignac, Delft3D User Days
2016
3
Image from CIPEL, 2016
Monitoring station SHL2
Introduction – Material and methods – Results and discussion – Conclusions

4. Satellite observations
• Chla heterogeneities in Lake Geneva (Kiefer et al, 2015)
Frédéric Soulignac, Delft3D User Days
2016
4
Do spatio-temporal heterogeneities bias our estimation of lake ecological
status?
More generally, do these heterogeneities induce uncertainties in our
classification of lakes by ecological quality?
Introduction – Material and methods – Results and discussion – Conclusions

5. Objectives
• To simulate 5 parameters used for the estimation of lake ecological status
(Chla, NH4, NO3, TP and Secchi)
• To evaluate the variability of Lake Geneva ecological status linked to the
choice of the 4 sample dates
• To evaluate the spatial variability of Lake Geneva ecological status caused
by spatial heterogeneities
• To locate the most representative sampling station of Lake Geneva
5
Frédéric Soulignac, Delft3D User Days
2016
Introduction – Material and methods – Results and discussion – Conclusions

6. Numerical domain
Frédéric Soulignac, Delft3D User Days
2016
6
820 curvilinear grids (resolution 1 km2)
100 vertical Z-layers (resolution from 1 to 7 m)
Introduction – Material and methods – Results and discussion – Conclusions

7. Delft3D-FLOW model
Frédéric Soulignac, Delft3D User Days
2016
7
A
B
C D
A
B
C
D
Introduction – Material and methods – Results and discussion – Conclusions
Wind episode
Mixing regime explains Chla heterogeneity (Soulignac et al., 2016)

8. Delft3D-ECO processes configuration
• Reminder of the 1st objective: to simulate Chla, NH4, NO3, TP and Secchi
• Primary producers: BLOOM (Los, 2009)
– Chla derived from 4 phytoplankton groups: diatoms, flagellates, green and blue green algae
• Dissolved inorganic matter
– 4 nutrients: NH4, NO3, PO4 and SiO2
– Opal and AAP
• Organic matter
– 4 fraction of particulate organic matter (POC1-4, PON1-4 and POP1-4)
– Dissolved organic carbon, nitrogen and phosphorus (DOC, DON and DOP)
• TP: dissolved and particulate phosphorus (including algae)
• Secchi: phytoplankton, particulate organic and inorganic matter
• Primary consumers
Frédéric Soulignac, Delft3D User Days
2016
8
Introduction – Material and methods – Results and discussion – Conclusions

9. Simulation realization
• 4 years (2009-2012), 1 simulation per year
• Initial condition (horizontally homogenous)
– Measurements at monitoring station SHL2
– NH4, NO3, PO4, SiO2, POC, PON, POP
• Hydrodynamic forcing (from Delft3D-FLOW model)
– Water velocity
– Water temperature
– Vertical dispersion
• Ecological forcing
– Meteorology
• Daily solar radiation
• Constant wind speed (reaeration)
– Tributaries: Weekly flow rates and composition (measurements)
– Zooplankton
Frédéric Soulignac, Delft3D User Days
2016
9
+
Monitoring station SHL2
Introduction – Material and methods – Results and discussion – Conclusions

10. • Sampling protocol for biological and physico-chemical parameters
(French technical guide for evaluating ecological status of lakes and
reservoirs, 2016 ; De Bortoli and Argillier, 2008)
– 4 integrated samples in the euphotic zone
– At 1 monitoring station, the deepest point of the lake
– Observations over the growing season, one per temporal window (TW1-4)
• Annual metrics
– Chla: mean value of TW2-4
– NH4 and NO3: maximum value of TW1-4
– TP and Secchi: median value of TW1-4
Calculation of lake ecological status in France
Frédéric Soulignac, Delft3D User Days
2016
10
Introduction – Material and methods – Results and discussion – Conclusions

11. Calculation of Lake Geneva ecological status
from Delft3D
Parameters Ecological status
Frédéric Soulignac, Delft3D User Days
2016
11
Introduction – Material and methods – Results and discussion – Conclusions

12. Localisation of the most representative
monitoring station
Ecological status Representativeness
Frédéric Soulignac, Delft3D User Days
2016
12
Introduction – Material and methods – Results and discussion – Conclusions

13. Ecological model performances
Frédéric Soulignac, Delft3D User Days
2016
13
Monitoring station SHL2
Depth Average 0-30 m
+
Introduction – Material and methods – Results and discussion – Conclusions

14. Variability of Lake Geneva ecological
status linked to sample dates
Frédéric Soulignac, Delft3D User Days
2016
14
+
Monitoring station SHL2
Year 2010
Chla NH4 NO3 TP Secchi
Bad 0 % 0 % 0 % 0 % 0 %
Poor 98 % 0 % 0 % 0 % 0 %
Medium 2 % 0 % 0 % 4 % 0 %
Good 0 % 0 % 100 % 89 % 51 %
Very good 0 % 100 % 0 % 7 % 49 %
Introduction – Material and methods – Results and discussion – Conclusions

15. Spatial variability of Lake Geneva ecological
status
15
Frédéric Soulignac, Delft3D User Days
2016
Year 2010
Introduction – Material and methods – Results and discussion – Conclusions

16. Most representative monitoring station (1/2)
16
Frédéric Soulignac, Delft3D User Days
2016
Year 2010
Introduction – Material and methods – Results and discussion – Conclusions

17. Most representative monitoring station (1/2)
17
Frédéric Soulignac, Delft3D User Days
2016
Year 2010
Most representative sampling station
Monitoring station SHL2
Introduction – Material and methods – Results and discussion – Conclusions

18. Conclusions
• A method is proposed to quantify the variability of lake ecological status and
locate the most representative sampling station by using Delft3D in
agreement with WFD
• Seasonal variability of Lake Geneva ecological parameters is well
reproduce by the model
• Spatio-temporal heterogeneities of ecological parameters in Lake Geneva
– induce a low variability on ecological status based on Chla
– don’t induce any variability on ecological status based on NH4 and NO3
– induce a higher variability on ecological status based on TP and Secchi
• Heterogeneities can bias our evaluation of lake ecological status but only for
some parameters, for the other parameters the proposed metrics to
evaluate an ecological status are robust and the trophic status estimated is
not influenced by spatio-temporal variability
18
Frédéric Soulignac, Delft3D User Days
2016
Introduction – Material and methods – Results and discussion – Conclusions

19. 1: The French National Institute for Agricultural Research - Inra
Lake Hydrobiology Unit, Thonon-les-Bains, France
Thank you!
Delft3D User Meeting - 03 November 2016
Frédéric Soulignac1, P.-A. Danis2,10, D. Bouffard3,
V. Chanudet4, E. Dambrine5, Y. Guenand6, B. Guillermin1,
T. Harmel7,10,2, B. Ibelings8, I. Kiefer3, D. Trevisan1,
R. Uittenbogaard9 and O. Anneville1
3 4 5 6 7 8 9 10
2: The French National Agency for Water and Aquatic Environments