Title: 3rd plenary meeting, March 67, 2006, London
1 3rd plenary meeting, March 6-7, 2006,
London MERSEA Task 7.2.2 Integrated Ecosystem
model in OPA Implementation of a coupled
physical biological model with data
assimilation Yann Ourmières, Pierre Brasseur,
Léo Berline, Jean-Michel Brankart and Jacques
Verron LEGI / CNRS Grenoble, France
2- Main goals for the 7.2 task
- Development of a state-of-the-art biogeochemical
model (NPZD global ecomodel, low resolution to
high resolution) - Implement and adapt a data assimilation
methodology into a coupled physical-biogeochemical
model (subtask 7.2.2) - Realize the necessary transfers for a
demonstration experiment at global scale and high
resolution (1/4) of a coupled physico-biogeochemi
cal assimilating model
3Context of the study
- Expertise in assimilation of physical data for
circulation models (SEEK filter developed at the
LEGI, transfered to MERCATOR SAM2) - ? Interest in understanding the impact of
physical data assimilation on the coupled system
and especially on the ecosystem response - On-line coupled model
- 3D Basin-scale configuration
- Use of physical data only for assimilation
-
4Main results after 12 months Beta version
developed and assessed
5Prototype of a coupled physical-biological
assimilative system
- Beta-version details
- OPA model rigid lid, eddy permitting
resolution (1/3) - OPA 8.1, North Atlantic domain (MNATL)
-
- First generation biogeochemical model NPZD
- - 1 nutrient (PO4)
- - coupled on line
- Sequential assimilation scheme (SEEK filter)
- - Singular Evolutive Extended Filter (Pham et
al., 1998) - - Assimilated data
- - Surface physical data only
- (TP/ERS altymetry, AVHRR
SST)
6Summary of the coupled beta-version
performance without assimilation
(year 1998)
7Annual maximum of the mixed layer depth (m) for
1998
Free run
Assimilated run
Climatology from in-situ data (De Boyer Montegut
et al.,2004)
8Impact on the ecosystem Spring mean of surface
Chl-a (mg/m3) for 1998
Assimilated run
Free run
underestimation
Gradient still too strong
Observations (SeaWifs)
mg/m3
L. Berline, J-M. Brankart, P. Brasseur, Y.
Ourmières, and J. Verron Improving the physics
of a coupled physical-biogeochemical model of the
North Atlantic through data assimilation impact
on the ecosystem, in revision for the Journal of
Marine Systems, special issue AMEMR symposium
june 2005
9- Problems highlighted
- Nutrient input strongly affected by data
assimilation, but its effect depends on the
region - Remaining flaws of the biological model response
(e.g. grazing at high latitudes, subtropical gyre
too oligotrophic...) - Discontinuous nature of the sequential
assimilation system leading to unrealistic
vertical advection - Consistent modification of the assimilation
system ? - By obtaining a continuous assimilated solution
(use of IAU as a first step to a SEEK smoother) - By performing more consistent corrections
between the physical and the biogeochemical
variables.
10Work progress after 18 months Platform upgraded
and assessment in progress
11Upgrade of the assimilated coupled model OPA9
LOBSTER - AS
- Physical model OPA9 in NATL4 configuration
- OPA9 bulk forcing (mld control, S.Skachko),
free surface, partial steps - NATL4 North Atlantic domain
- ¼ resolution
- 45 vertical level, thickness of
6m (surf.) to 250m (bottom) - Biogeochemical model LOBSTER
- N2PZD2 model of intermediate complexity
- 2 nutrients
- Coupled on-line to the physical model
- Assimilation system (AS) SEEK IAU
- Implementation of IAU filtering to obtain a
continuous solution and to reduce analysis
induced shocks. - Assimilated data
- - Surface physical data only
- (TP/ERS altymetry, AVHRR SST)
12First results of the upgraded coupled system
without assimilation
SeaWifs monthly climatological Chlorophyll a for
May
Chlorophyll a (mg/m3) from a model snapshot. Date
29/05/96. No assimilation
13 Work plan for subtask 7.2.2
- 1) Evaluation of the performance of the coupled
system, focusing on the biogeochemical component
response - Assessment based on metrics developed with
beta-version and WP 7.2.3. - Modification / enhancement of assimilation system
for key coupling mechanism (vertical advection,
mixed layer dynamics...)
2) 2 years simulation (98-99) and validation
with data such as ocean colour products
(SeaWifs) 3) Transfer for demonstration with the
MERCATOR system
14WP 7.2.2 Schedule
0 6 12 18 24
30 36 42
48 04/04 10/04 04/05
10/05 04/06 10/06 04/07
10/07 04/08
15Research directions for the assimilative system
- Use of ocean colour data for a better parameter
estimation - Assimilation method using flux correction
(presentation from S. Skachko) - Combining surface and in-situ data for
assimilation and/or validation