PPR MERCATORVERT: Putting Bio into Operational Oceanography

1
PPR MERCATOR-VERT Putting Bio into
Operational Oceanography

• Coordination C. Moulin and P. Monfray
• towards GMES/MCS
• towards a joint GODAE-IMBER joint WG

2
Objectives

• Use GREEN MERCATOR as a brainstorming network and
  training on Bio into OO
• Exchange ideas for a virtuous loop between OO and
  research
• Use meetings with a lot of time for informal
  discussions

3
BIONUTS (2004-2006)
RD
Implementation
WP1 (P. Monfray) N.Atlantic Demonstrator PSY1-LOB
STER PSY1-PISCES
WP2 (P. Brasseur) Assimilation
P3ZD to LOBSTER
4
The BIONUTS heritage
Chlorophylle de surface pour le mois de Mars 1998
SeaWiFS
MERA11-v067

• Improved assimilation techniques of physical
  parameters, but still weaknesses
• Increase the resolution?
• Assimilate other parameters? (NO3, Chl,)

utilisant le version PSY1V2 de MERCATOR
5
The BIONUTS heritage
Putting BIO into OO is doable !!
Implémentation du démonstrateur en temps réel
LOBSTER-MERCATOR
Applications Regional/Coastal OO, higher trophic
levels,
6
Green Mercator (2007-2009)
Application
Implementation
RD
WP4 (O. Aumont) Interface to regional sub-system
WP1 (M.Levy) Improvement of process resolution
PSY3-PISCES WP3 (C. Moulin) Extension to
global ocean PSY4-PISCES
WP2 (P. Brasseur) Model-data integration
WP5 (P. Lehodey) Extension towards marine
resources
Year 1 2
Year 2 3
WP6 Evaluation strategy (I. Allen)
Year 3
7
Le Projet

• Kick-off meeting in March
• 2007 in Paris
• 57 k pour 3 ans
• Retard dans la mise en place des crédits lié à la
  signature de la convention entre le CEA, le CNRS
  et le GIP Mercator.

8
WP1 improvement of process resolution
Based on analysis and reanalysis at various
resolutions to investigate the impact on
biogeochemistry of the transition between eddy
permitting to eddy resolving Task 1.1
Evaluation of space-time resolution Task 1.2
Impact of resolution over the Indian Ocean
M. Lévy, O. Aumont, A. Albert (stage M2), D.
Iovino (post-doc), V. Koné (post-doc), C.
Lathuillière (thèse), L. Resplandy (thèse)
9
Task 1.1 Evaluation of space-time resolution
10
Task 1.1 Evaluation of space-time resolution
11
Task 1.1 Evaluation of space-time resolution
B9 online
B9 off line

• How to represent
• vertical mixing
• in offline run with
• degraded resolution?

Phyto Surface Moyenne annuelle
B27 online
B27onB9 offline
12
Task 1.2 Impact of resolution over the Indian
Ocean
EFFECTUÉ 1 - ORCA05PISCES analyse du cycle
saisonnier et colimitations (Koné et al., in
prep) 2 - mise en place dun modèle PISCES
simplifié pour lOI (3 fois moins de traceurs) 3
- test DRAKKAR Pisces réduit en offline gros
pb sur les W aux maldives EN COURS 4 - mise en
place dune configuration frontières ouvertes
1/4 ocean Indien 5 - mise en place dune config
1/4 mer dArabie PREVU 6 - 1/4 Indien
PISCES réduit comparaison avec ORCA05-PISCES 7
- 1/4 mer dArabie PISCES réduit 8 - passage
au 1/12 sur mer dArabie
13
WP2 model-data integration
Based on assimilation of physical and/or
biogeochemical data to improve physics and/or
biogeochemistry for biological purposes Task 2.1
development of a ¼ North Atlantic prototype,
assimilating physical AND nutrient data in a
coupled model Task 2.2 Biological model
parameter estimation via ocean colour data
assimilation
D. Béal, J.-M. Brankart, P. Brasseur, Y.
Ourmières(), J. Verron LEGI/CNRS, Grenoble (
now at LSEET/USTV, Toulon) Kane, C.
Moulin IPSL/LSCE, Gif-sur-Yvette
14
Task 2.1 Assimilative Platform (LEGI)
North Atlantic configuration Domain 20S-80N,
98W-23E Resolution 1/4 - 21 km in the Gulf
Stream 45 levels, 6m to 250 m Free
surface formulation, partial steps Surface
fluxes ERA40 (bulk formulations) LOBSTER coupled
ON-LINE! Experimental set-up Spin-up period
interannual 1984-1996 Hindcast period
1996-1998 Assimilation scheme SEEK filter,
local error sub-space (EOFs), incremental
updates Assimilation data along-track T/P ERS
altimetry, AVHRR SST and climatological nutrient
data (Levitus)
15
May 1998 surface chlorophyll (mg.m-3)
FREE
SeaWiFS
PH_A
PN_A
16
June 1998 Nitrate (mmolN.m-3)
FREE
Levitus
PH_A
PN_A
17
Task 2.1 Assimilative Platform (LEGI)

• The assimilative platform works well
• Demonstrated utility to constrain not only the
  physics but also the components of the biological
  state vector
• The tools to assimilate nitrate data O.K., but we
  need reliable in situ data !
• Ocean colour should be assimilated too
  observability ?
• (Details in Ourmières et al., 2007, submitted)

18
Task 2.2 Towards Ocean Colour Assimilation
(LEGI)
Context Data Assimilation within a coupled
physical biogeochemical model Objective
ensemble simulations to characterize the
sensitivity of the coupled model to error
sources (e.g. wind), in the perspective of
ocean colour data assimilation ?
what can be controlled with ocean colour data ? ?
? Strategy Before data assimilation (DA)
(i) to explore the impact of wind errors in the
physical model on biogeochemical
variables, using ensemble techniques (ii)
then, to investigate the inverse problem for DA.
19
Task 2.2 Towards Ocean Colour Assimilation
(LSCE)
Context Biogeochemical data assimilation within
PISCES Objective Optimize PISCES physiological
parameters (mortality, grazing, growth), in the
perspective of ocean colour data assimilation
Strategy Work on a 1D version of PISCES and
assume the physical model is OK
Chl at BATS
20
Task 2.2 Towards Ocean Colour Assimilation
(LSCE)
21
WP3 extension to global ocean
Based on analysis and reanalysis at 1/4
(eventually at 1/12), to investigate
variability of primary production and CO2
fluxes Task 3.1 Integration of degradation
into the Mercator system Task 3.2
Implementation of PISCES in a global version

• C. Moulin, L. Bopp
• LSCE
• O. Aumont
• LOCEAN
• A. El Moussaoui, M. Drévillon, E. Dombrowsky, L.
  Nouel, E. Greiner
• MERCATOR

22
Task 3.1 Integration of degradation
La dégradation est une opération qui consiste à
moyenner la dynamique et les forçages sur les
différents points de grille mer qui
correspondent à un super-point mer de la
grille grossière.
23
Task 3.1 3.2 Integration of Degradation and
of PISCES
NATL4 (1/4)
Achevé (quasi)
. PISCES/BIO4
Spin-up 3 ans
Simulations 2002-2005
. PISCES/BIO1 , dégradé de 1/4 ? 1
Spin-up 5 ans
Simulations 2002-2005
Globale ¼
En cours
. Version dégradée 1/4 ? 1
Spin-up 10 ans
Simulations 2002-2005
Prévu 2008
NATL 1/12
24
Task 3.1 3.2 Integration of Degradation and
PISCES
25
WP4 Interface to regional sub-system
Based on interface coupler for physical and
biogeochemical variables, allowing the use of
global ocean analysis or reanalysis to force
coastal or regional biogeochemical/ecosystem
models

• O. Aumont, V. Echevin, V. Kone, M. Lévy, F.
  Pinsard
• LOCEAN
• D. Obaton
• MERCATOR

26
WP4 Interface to regional sub-system

• Objectives
• Set up a regional configuration at high
  resolution with ROMS/PISCES
• Develop the tools to extract the forcing fields
  and the boundary conditions
• Automate the system
• Set up a regional operational demonstrator that
  runs automatically after MERCATOR
• Deliverables
• 1st year Set up the regional configuration,
  validate it and develop the preprocessing tools
• 2nd year Port the system to the MERCATOR
  facilities, use the output from WP3
• 3rd year Automation of the system (demonstrator)
  and distribution of the results

27
WP4 Interface to regional sub-system

• Peru upwelling, 1/6, Climatological

SeaWiFS
Model
Data
Annual mean surface Chla (mg Chl/m3)

• Peru Upwelling
• This configuration is being used in several other
  projects
• Numerous in situ data, both physical and
  biogeochemical
• 1/12 is too expensive for the development

28
WP4 Interface to regional sub-system

• Achievements
• ROMS/PISCES code has been strongly optimised
• MPP capabilities (OpenMP, MPI) for tracers have
  been coded and validated
• Diagnostics for PISCES in ROMS have been added
• Pre-processing tools have been developped, both
  for climatological and ORCA05 runs
• Interannual simulations will be launched before
  the end of the year
• The forecast system works for the physics
  (biogeochemistry !!!)
• AGRIF for tracers is currently being tested

29
WP5 Extension towards marine resources
Based on offline coupling with medium, and
eventually upper, tropic level models
P. Lehodey, I. Senina, P. Gaspard CLS O.
Aumont LOCEAN
30
SODA
1ère phase réalisée

• PP et Zeu dérivés des données satellites

• En labsence de réanalyse MERCATOR, la réanalyse
  SODA est utilisée.

• Simulation globale à la résolution ½ deg x mois
  pour la période sep 1997-déc 2004.

31
Comparison of primary production (above) derived
from satellite (Behrenfeld et al., 2006) and
predicted distribution of daytime epipelagic
micronekton biomass (Jan 2004)
32
Mercator VertWP6 Evaluation Strategy Task 6.1
Promoting a set of metrics to benchmarks model
performance

• I. Allen
• Plymouth Marine Laboratory

33
Ordre du jour de la Réunion du projet

• 14h-15h30 PRESENTATIONS DE L'AVANCEMENT DES
  TRAVAUX
• Chaque WP leader présente l'avancement des
  travaux, les difficulté rencontrées et les
  travaux prévus de FACON SYNTHETIQUE en quelques
  transparents.
• 15h30-17h30 DISCUSSIONS
• INTERACTIONS ENTRE WP c'est la spécificité du
  projet, chaque WP etant organisé autour de
  projets propres. Des échanges sont établis entre
  WP1 et WP2 et entre WP1 et WP3 (dégradation/résolu
  tion), mais il serait bien de se fixer des
  objectifs communs. Les WP4 et WP5 semblent plus
  "isolés" et se positionnent en utilisateurs des
  simulations du WP3, il faut aussi réfléchir a une
  façon de renforcer ses liens. Un objectif
  important va être de planifier quelques réunions
  en comité réduit pour faire avancer ces
  collaborations.
• POINTS SUR LES PROJETS ASSOCIES en particulier
  sur les projets FP7acceptés ou en cours
  d'acceptation (My_Ocean, MEECE, HiSEA,...), mais
  aussi sur les projets nationaux. Ce serait bien
  d'avoir un ou deux transparents sur chacun de ces
  projets pour que tout le monde ait une vision
  complète de la façon dont ces projets vont
  contribuer à Mercator-Vert (CDD, postdoc,
  doctorants, fonctionnement).
• POINTS SUR LES DIFFICULTES RENCONTREES en
  particulier en termes technique (degradation,
  ORCA025, couplage NATL4-BIO1,...). Un objectif
  important va être de planifier quelques réunions
  en comité réduit pour résoudre efficacement ces
  difficultés.
• QUESTIONS DIVERSES Besoin de chaque WP en
  termes de calcul sur lIDRIS, site web,

34
Task 2.1 Assimilative Platform (LEGI)

• The assimilative platform works well
• Demonstrated utility to constrain not only the
  physics but also the components of the biological
  state vector
• The tools to assimilate nitrate data O.K., but we
  need reliable in situ data !
• Ocean colour should be assimilated too
  observability ?
• (Details in Ourmières et al., 2007, submitted)

35
Task 2.2 Towards Ocean Colour Assimilation
(LEGI)
Context Data Assimilation within a coupled
physical biogeochemical model Objective
ensemble simulations to characterize the
sensitivity of the coupled model to error
sources (e.g. wind), in the perspective of
ocean colour data assimilation ?
what can be controlled with ocean colour data ? ?
? Strategy Before data assimilation (DA)
(i) to explore the impact of wind errors in the
physical model on biogeochemical
variables, using ensemble techniques (ii)
then, to investigate the inverse problem for DA.
36
Task 2.2 Towards Ocean Colour Assimilation
(LSCE)
Context Biogeochemical data assimilation within
PISCES Objective Optimize PISCES physiological
parameters (mortality, grazing, growth), in the
perspective of ocean colour data assimilation
Strategy Work on a 1D version of PISCES and
assume the physical model is OK
Chl at BATS
37
Task 2.2 Towards Ocean Colour Assimilation
(LSCE)
38
WP3 extension to global ocean
Based on analysis and reanalysis at 1/4
(eventually at 1/12), to investigate
variability of primary production and CO2
fluxes Task 3.1 Integration of degradation
into the Mercator system Task 3.2
Implementation of PISCES in a global version

• C. Moulin, L. Bopp
• LSCE
• O. Aumont
• LOCEAN
• A. El Moussaoui, M. Drévillon, E. Dombrowsky, L.
  Nouel, E. Greiner
• MERCATOR

39
Task 3.1 Integration of degradation
La dégradation est une opération qui consiste à
moyenner la dynamique et les forçages sur les
différents points de grille mer qui
correspondent à un super-point mer de la
grille grossière.
40
Task 3.1 3.2 Integration of Degradation and
of PISCES
NATL4 (1/4)
Achevé (quasi)
. PISCES/BIO4
Spin-up 3 ans
Simulations 2002-2005
. PISCES/BIO1 , dégradé de 1/4 ? 1
Spin-up 5 ans
Simulations 2002-2005
Globale ¼
En cours
. Version dégradée 1/4 ? 1
Spin-up 10 ans
Simulations 2002-2005
Prévu 2008
NATL 1/12
41
Task 3.1 3.2 Integration of Degradation and
PISCES
42
WP4 Interface to regional sub-system
Based on interface coupler for physical and
biogeochemical variables, allowing the use of
global ocean analysis or reanalysis to force
coastal or regional biogeochemical/ecosystem
models

• O. Aumont, V. Echevin, V. Kone, M. Lévy, F.
  Pinsard
• LOCEAN
• D. Obaton
• MERCATOR

43
WP4 Interface to regional sub-system

• Objectives
• Set up a regional configuration at high
  resolution with ROMS/PISCES
• Develop the tools to extract the forcing fields
  and the boundary conditions
• Automate the system
• Set up a regional operational demonstrator that
  runs automatically after MERCATOR
• Deliverables
• 1st year Set up the regional configuration,
  validate it and develop the preprocessing tools
• 2nd year Port the system to the MERCATOR
  facilities, use the output from WP3
• 3rd year Automation of the system (demonstrator)
  and distribution of the results

44
WP4 Interface to regional sub-system

• Peru upwelling, 1/6, Climatological

SeaWiFS
Model
Data
Annual mean surface Chla (mg Chl/m3)

• Peru Upwelling
• This configuration is being used in several other
  projects
• Numerous in situ data, both physical and
  biogeochemical
• 1/12 is too expensive for the development

45
WP4 Interface to regional sub-system

• Achievements
• ROMS/PISCES code has been strongly optimised
• MPP capabilities (OpenMP, MPI) for tracers have
  been coded and validated
• Diagnostics for PISCES in ROMS have been added
• Pre-processing tools have been developped, both
  for climatological and ORCA05 runs
• Interannual simulations will be launched before
  the end of the year
• The forecast system works for the physics
  (biogeochemistry !!!)
• AGRIF for tracers is currently being tested

46
WP5 Extension towards marine resources
Based on offline coupling with medium, and
eventually upper, tropic level models
P. Lehodey, I. Senina, P. Gaspard CLS O.
Aumont LOCEAN
47
SODA
1ère phase réalisée

• PP et Zeu dérivés des données satellites

• En labsence de réanalyse MERCATOR, la réanalyse
  SODA est utilisée.

• Simulation globale à la résolution ½ deg x mois
  pour la période sep 1997-déc 2004.

48
Comparison of primary production (above) derived
from satellite (Behrenfeld et al., 2006) and
predicted distribution of daytime epipelagic
micronekton biomass (Jan 2004)
49
Mercator VertWP6 Evaluation Strategy Task 6.1
Promoting a set of metrics to benchmarks model
performance

• I. Allen
• Plymouth Marine Laboratory

50
Ordre du jour de la Réunion du projet

• 14h-15h30 PRESENTATIONS DE L'AVANCEMENT DES
  TRAVAUX
• Chaque WP leader présente l'avancement des
  travaux, les difficulté rencontrées et les
  travaux prévus de FACON SYNTHETIQUE en quelques
  transparents.
• 15h30-17h30 DISCUSSIONS
• INTERACTIONS ENTRE WP c'est la spécificité du
  projet, chaque WP etant organisé autour de
  projets propres. Des échanges sont établis entre
  WP1 et WP2 et entre WP1 et WP3 (dégradation/résolu
  tion), mais il serait bien de se fixer des
  objectifs communs. Les WP4 et WP5 semblent plus
  "isolés" et se positionnent en utilisateurs des
  simulations du WP3, il faut aussi réfléchir a une
  façon de renforcer ses liens. Un objectif
  important va être de planifier quelques réunions
  en comité réduit pour faire avancer ces
  collaborations.
• POINTS SUR LES PROJETS ASSOCIES en particulier
  sur les projets FP7acceptés ou en cours
  d'acceptation (My_Ocean, MEECE, HiSEA,...), mais
  aussi sur les projets nationaux. Ce serait bien
  d'avoir un ou deux transparents sur chacun de ces
  projets pour que tout le monde ait une vision
  complète de la façon dont ces projets vont
  contribuer à Mercator-Vert (CDD, postdoc,
  doctorants, fonctionnement).
• POINTS SUR LES DIFFICULTES RENCONTREES en
  particulier en termes technique (degradation,
  ORCA025, couplage NATL4-BIO1,...). Un objectif
  important va être de planifier quelques réunions
  en comité réduit pour résoudre efficacement ces
  difficultés.
• QUESTIONS DIVERSES Besoin de chaque WP en
  termes de calcul sur lIDRIS, site web,