Ocean Ecosystem Modeling and Observations

1
Ocean Ecosystem Modeling and Observations

• IARC Contributors
• Ana Aguilar-Islas
• David Atkinson
• Clara Deal
• Meibing Jin
• Peter McRoy
• Eiji Watanabe
• Jingfeng Wu

Collaborating Institutions include Antarctic
Climate and Ecosystems Cooperative Research
Centre, Tasmania, Australia Genwest Systems,
Seattle, Washington, USA Graduate School of
Fisheries, Hokkaido University, Hokodate,
Japan Korean Polar Research Institute, Republic
of Korea Los Alamos National Laboratory (LANL),
Los Alamos, New Mexico, USA NOAA, Great Lakes
Regional Laboratory, Ann Arbor, Michigan,
USA School of Fisheries and Ocean Sciences and
Institute of Arctic Biology, UAF, Alaska,
USA University of Groningen, The Netherlands
2
Block diagram outlining some of our strengths and
how we work together.
Ocean Ecosystem Modeling and Observations Theme
Observations
Modeling
Fe biogeochemistry Dimethylsulfide (DMS)
cycle Terrestrial C input and fate CO2 and
methane dynamics
Satellite remote sensing Internet
databases Reanalysis data
Statistical GIS-based model Surface seawater DMS
Process-based models 1-D Physical ice-ocean
ecosystem model DMS cycling Fe limitation on
productivity Inorganic C component 3-D Physical
ice ocean ecosystem model 3-D Eddy-resolving ice
ocean model Working towards Artic System Model
(ASM)
3
(No Transcript)
4
Diagram illustrates integration within ocean
ecosystem modeling theme and among 2nd stage
themes, and what feedbacks to the physical system
might be expected.
Role of Freshwater/Permafrost in the
Arctic-global Connection
Global effect of Warming in the Arctic
Radiation Budget
Increased Precipitation Warming Permafrost
CCN
-
Fate of Sea Ice in the Arctic Ocean
Sulfate aerosols
Global Temperature
Increased Discharge
Radiative Feedbacks?
SO2
Decreased Ice/ Increased Open Water
Radiative Feedbacks?
CO2
Runoff, C, nutrients
DMS
atmosphere
CO2
CH4
?
?
Erosion C, nutrients
?
ice
ice
CO2
Dissolved inorganic carbon (DIC)
CO2
CH4
marine food web
Increased Erosion
Fe
Nutrients
Increased Storminess, Diminished Sea Ice
Warming Permafrost
euphotic zone
DOC
POC
CaCO3
Export
DOC
POC
CaCO3
DIC
deep ocean
5
The incorporation of iron into marine ecosystem
models has just begun in recent years.
Questions How and in what form delivered? How
made available to phytoplankton? How cycled in
marine ecosystem?
CCN
Radiation Budget
Sulfate aerosols
-
SO2
Radiative Feedbacks?
Global Temperature
DMS
Radiative Feedbacks?
CO2
Runoff, C, nutrients
?
atmosphere
CO2
CH4
?
Erosion C, nutrients
?
ice
ice
CO2
CO2
Dissolved inorganic carbon (DIC)
CH4
marine food web
Fe
Nutrients
euphotic zone
DOC
POC
CaCO3
Export
DOC
POC
CaCO3
DIC
deep ocean
6
Sources of iron to the surface ocean.
Bering SeaApril
Calvin Mordy unpublished data
green, 050 m yellow, 50125 m red, 125200
m gray 200275 m blue, 275 m
Beaufort Sea May to November
Simpson et al., 2008
7
Sea ice-derived dissolved iron influences the
spring algal bloom in the outer shelf and shelf
break of the Bering Sea.
(Aguilar-Islas, Wu)
Fe (nM)

• In the mid and inner shelf sedimentary iron
  inputs can reach surface waters during spring.

• In the outer shelf and shelf break melting sea
  ice provides additional iron for the complete
  assimilation of available nitrate by large
  cells. Aguilar-Islas, Wu, et al. , 2008 (GRL)

8
Convergence of nitrate-rich offshore waters with
iron-rich coastal waters leads to high
productivity in the NW Gulf of Alaska.
(Wu, Aguilar-Islas)
GAK 1
GAK 7
GAK 1
GAK 13
Continental input rich in Fe
GAK 7
GAK 13
Wu et al. (submitted, GRL)
9
Aerosol Fe solubility dominated by the colloidal
fraction.
Estimates of aerosol iron solubility in seawater
reported in the literature (0.01-90)
(Aguilar-Islas, Wu)
Fe dissolution
(0.02 µm lt colloidal lt 0.4 µm)
Different leaching solutions
Aerosols collected from different areas
(dissolved is lt 0.4 µm)
Most of the aerosol iron dissolved in seawater
was in the colloidal size fraction
Aguilar-Islas, Wu, et al. , 2009 (Marine
Chemistry)
Aguilar-Islas et al. In press (Marine Chemistry)
10
Diminishing arctic sea ice will influence
important biogeochemical cycles and the marine
ecosystem.
CCN
Radiation Budget
Sulfate aerosols
-
SO2
Radiative Feedbacks?
Radiative Feedbacks?
Global Temperature
DMS
CO2
Runoff, C, nutrients
?
atmosphere
CO2
CO2
CH4
?
Erosion C, nutrients
?
ice
ice
CO2
CO2
CH4
Dissolved inorganic carbon (DIC)
marine food web
CO2
Fe
euphotic zone
DOC
POC
CaCO3
Export
DOC
POC
CaCO3
DIC
deep ocean
11
On the relative importance of the functional
relationships and feedbacks to a Pan-Arctic
perspective and thus to a comprehensive ASM.

• Feedbacks that involve clouds are particularly
  relevant to the Arctic because clouds influence
  the physical processes most important to the
  warming of the Arctic and the melting of sea ice.
• Clouds remain one of the largest uncertainties in
  climate modeling.
• Cloud properties such as albedo, extent, and
  duration are determined in large part by cloud
  condensation nuclei (CCN).
• Source of CCN over the summertime Arctic is
  nucleated particles of marine biogenic origin
  that grow to CCN size with the aid of aerosol
  precursor gases, predominantly DMS.

Kettle et al. (1999) DMS climatology updated by
Belviso et al. (2004).
Late-spring low stratus offshore Barrow, Alaska.
12
What is the impact of DMS on climate?

• Recent climate models (Gunson et al. 2006)- 50
  reduction of ocean DMS emission radiative
  forcing 3 W/m2 air temperature 1.6 C-
  doubling of ocean DMS emission radiative
  forcing -2 W/m2 air temperature -0.9 C
• Model projections (Gabric et al. 2004) impact of
  warming on the global zonal DMS flux (70 N- 70 S)
  indicates greatest perturbations to be at high
  latitudes
• Use of a climate model to force ocean DMS model
  in Barents Sea (Gabric et al. 2005)
• - By the time of equivalent CO2 tripling
  (2080)
• zonal annual DMS flux increase gt80
• zonal radiative forcing -7.4 W/m2
• summer (June-September)

13
Recent DMS modeling and observations.
A statistical (GIS-based) approach.
(M.S. graduate student Humphries, Deal, Atkinson)
Modeled surface DMS for month of May.
(Deal, Jin)
Influence of sea ice on marine sulfur
biogeochemistry in Community Climate System Model
(CCSM), July 2009-2012.
Field and laboratory studies help to clarify
sub-processes.
(DMS in sea ice, J. Stefels, unpublished data)
14
Ecosystem modeling focus on ice-ocean ecosystem.
IARC ocean DMS ecosystem model (Jodwalis (Deal)
et al. 2001)
IARC ice-ocean ecosystem model applied Land-fast
ice in Chukchi Sea (Jin, Deal et al.
2006) Fluctuating ice zone of Bering Sea (Jin,
Deal et al. 2007 Jin, Deal, McRoy et al.
2008) Multi-year pack ice Canadian Basin (Lee,
Jin, et al. submitted)
Working towards a more regenerative
microbial loop in ice ecosystem model,
CO2
Dissolved inorganic carbon (DIC)
marine food web
Fe
and, Fe limitation on phytoplankton growth.
15
Model results show phytoplankton bloom patterns
in the southeastern Bering Sea are related to the
Pacific Decadal Oscillation (PDO) Index regimes.
Comparison of modeled phytoplankton at the
southeastern Bering Sea with a) daily SeaWiFS
data at sea surface b) mooring fluorescence
data at 12 m.
(Jin, Deal, McRoy)
Modeled monthly mean net primary
production (NPP) for years of PDO Index gt 1
subtracted by the mean for years of PDO Index lt
-1. D, F, Ai denote diatoms, flagellates, and
ice algae, respectively. Jin, Deal, McRoy, et
al. 2008 (JGR).
16
By implementing the IARC 1-D ice ecosystem model
in the LANL sea ice model, CICE, we have begun to
extend its scale.
(Deal, Jin)
Polar map of base ten logarithm mean ice bottom
layer Chl a concentration (mg Chl a m-2) for
mid-May. The white line is the 15 ice edge
contour and the black lines are ice thickness
contours of 1, 2, 3 and 4 m, working inward from
the ice edge.
(Jin, Deal)
Ice concentration (left) and ice algal biomass
(right) at the bottom of sea ice on May 13, 1981.
17
Pacific water transport from the Chukchi shelf to
the Canada Basin in being investigated with an
eddy-resolving coupled sea ice-ocean model.
(Watanabe AOMIP participant)
Model bathymetry m. B.S. Bering Strait, H.C.
Herald Canyon, C.C. Central Channel, N.R.
Northwind Ridge.
Center for Climate System Research Ocean
Component Model (COCO), Ver. 3.4, Univ. of
Tokyo 2.5 km horizontal explicitly resolves
mesoscale baroclinic eddies
Simulated eddies in vicinity of the Barrow
Canyon. Northward velocity averaged in the top
100 m in August is shaded cm s-1. Vectors
show ocean velocity averaged in the top 100 m and
their unit vector is 50 cm s-1.
18
The distribution of virtual tracer associated
with the Pacific water demonstrates that a
significant part of the Pacific water passes
through the Barrow Canyon during summer.
(Watanabe)
Seasonal cycle of transport of the virtual
Pacific water tracer across the dashed line Sv.
Vertically integrated concentration of virtual
Pacific water tracer in October m.
19
Plans - IARC Cooperative Agreement
Scientific Goal Quantify the relative current
and possible future influences of
arctic marine ecosystems on the global climate
system.
Hypotheses Enhanced DMS emissions from a more
ice-free Arctic Ocean will increase cloud
reflectivity of incoming solar radiation and
counter the initial loss of surface albedo
associated with the loss of sea ice. Changes in
arctic marine carbon cycle in response to a
warming climate will significantly influence
atmospheric CO2 and CH4 levels.
Strategy 1 Interface with modeling groups and
the other 2nd stage themes. Retrospective
studies with modeling tools. Experiments with
climate model runs for standard IPCC emissions
scenarios. Coupled model experiments with
GCMs. Provide ice ecosystem ocean ecosystem
module to ASM.
20
Challenges and Concerns
More in-situ and sustained observations needed.
http//saga.pmel.noaa.gov/dms/ Candian BioChem
database on web. Pan-Arctic PP database by P.
Matrai on web. Aguilar-Islas to study Fe in land
fast ice. Sea ice biologist R. Gradinger on IARC
team. Hydrological Atlas of the Bering Sea
Luchin Panteleev.
Verification of the 3-D physical-ecosystem models
and application in all critical areas.
Focus on the Bering-Chukchi-Beaufort Seas
Region. Group regions with similar features. Take
more advantage of in-house expertise.
Funds for students and post-docs.
New modeling and observations post-docs. Need to
work harder to recruit students.