Transcript E. Svendsen - Arctic Regional Ocean Observing System

Arctic Operational Oceanography
at
IMR
Einar Svendsen
Arctic GOOS planning meeting, 12-13 September 2006 at NERSC,
Bergen
An approach to marine ecosystem research and management
advice (with respect to science) is:
To consider the most important driving forces on,
and the processes within the ecosystems
Driving forces:
•Climate-physics (directly on all trophic levels and indirectly bottom-up
through the lower trophic levels
•Fisherman-fisheries management (top-down)
•Fertilization?
•Pollution?
•Introduction of new species?
•Habitat disturbance?
Climate-physics
Fishing
Climate-physics
Climate
Observations (from
ships satellites and
buoys) are important
for validation of and
assimilation into
the models
Svalbard
The
Nordic
seas
The
Barents
sea
Russia
Iceland
GB
This is our
theater
We are operational in giving fisheries advice (a few times a year),
but we are not operational in the “ecosystem approach” due to
lack of relevant oceanographic information and lack of
8
8
knowledge on how to use “operational oceanography”
Hindcast (50 year), nowcast and
forecast (week (or 100 years)) of:
Relevant physics
- Circulation, temperature, salinity, turbulence
Phytoplankton
- Concentration of functional groups (or specific (harmful) species),
nutrients, detritus, oxygen, sedimentation, light
Zooplankton
- Individual species (or functional group(s)? (IBM or Eulerian)
Fish larvae
- growth and distribution (and mortality?) (IBM)
Fish migration
- growth and distribution (overlap between species)
The operational needs
From the above variables, only physics is operationally available in hindcast,
nowcast and forecast (and still the quality can be questioned, partly due to lack of resolution due to
lack of computer resources.
Phytoplankton is starting to be operational (eg. MONCOZE, Liverpool Bay….)
We need zooplankton to realistically model larval growth and planktivour fish
migration, because this we need to more realistically address the key
challenges for the fisheries research, namely quantifying and predicting:
Recruitment, growth, mortality and distribution
Since we (mathematically) do not know all the processes leading up to these
states/processes, we need to make statistically shortcuts between smart
INDICATORS (derived from our modelled state variables) and recruitment,
growth, mortality and distribution, including observations where necessary.
NB! Overlap between pray and predators determines natural motality
Ideas for new operational indicators
- Position of fronts
- Extent and area of melting sea ice (if relevant)
- Area and volume of specific water masses
- Upwelling indexes
- Currents, temperature and turbulence
- Particle and tracer distributions from given sites (spawning, oil
production….)
- Fluxes of water masses and nutrients (though given sections)
- Timing (of peak spring bloom) and strength of primary prod.
- Light in water column
- Transport, growth and distribution of zoo-plankton
- Transport, growth and distribution of selected fish larvae
- Contaminant exposure on plankton and benthic ecosystems
- Sedimentation (resuspension)
- Overlap between species (prey and predators)
140
3.2
120
3
100
Fangst(Catch
Catch)
Vanntransport
2.8
80
2.6
60
2.4
40
2.2
20
2
0
-20
1975
1.8
1.6
1980
1985
1990
1995
2000
2005
Volumtransport (Sv)
Fangst (1000 tonn)
Predicting horse mackerel fishing from
modeled volume transport
2005 was extremely warm
Results: Monthly mean SST
Observed Climatology March
(Pathfinder)
Model results
March, 2004
50 year simulations demonstrate large decadal
variability
High NAO (1990-1995)
Low NAO (1963-1966)
We need to be continuously updated on the status
Variability in BSO-inflow
Anomalies from 1981-2000 average
Spawning and nursery grounds
Predators
Svalbard
Cod larvae
and early juveniles
Copepods
Russia
OCEAN CLIMATE
PARAMETERS
Transport
Temperature
Light conditions
Turbulence
Phytoplankton
Trophic transfer
Cod recruitment, volume transport and prim.prod.
Correlation map between primary production in
April and Cod (3Y) recruitment 3 years later
way
Nor
land
Green
32
26
32
26
Svendsen (2006)
Cod recruitment (3y) prediction
and ICES estimates
1200
Recruitment (million)
1000
800
600
400
200
0
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
Year
Priorities (conclusion)
• Finalize and analyse long term simulations (50 year) of
physics and phytoplankton globally, regionally and
locally. Easy access to results
• Develop and prepare operational systems ala
MONCOZE-MONBASE
• Implimentert realistic zooplankton modelling
operational and long term simulations
• Improve and implement fish growth and migration
models to explain the dynamics in natural mortality
• Couple models for pollution and biology to estimate
contaminant dozes on “stocks”
• Build systems where “bottom-up ecosystem approach”
can be useful for management advice on the marine
ecosystems
• Simulate potential ecosystem effects from climate
change
Climate