Decision support systems for ocean ecosystemclimate interactions

1
Decision support systems for ocean
ecosystem-climate interactions
Francisco Chavez Monterey Bay Aquarium Research
Institute
and mostly others
2
Conclusions

• Sophisticated observing systems, increased
  computing power for modeling and improved
  ecosystem theory provide the framework today to
  develop ecological forecasts useful for society
• Nature will continue to surprise (at a faster
  rate due to global change) and our theory will
  need to constantly evolve
• NASA will need to develop new direct or indirect
  estimates (and models) of ecologically relevant
  properties beyond chlorophyll
• Better working relationships between resource
  managers and NASA should be fostered

3
Roadmap

• How did we get here
• Approach
• Case Studies
• The way forward

4
The CUEA proposal funded by NSF in the 70s
said The goal of the Coastal Upwelling
Ecosystems Analysis Program is to understand the
coastal upwelling ecosystem well enough to
predict its response far enough in advance to be
useful to mankind. in the management of living
marine resources.
During the 1970s Bob Smith, Dick Dugdale and Dick
Barber believed that, in coastal upwelling
research, physics and biology together would be
more than the sum of the parts and together could
deliver an applied science product regarding
management of living marine resources. CUEA was
successful as an interdisciplinary basic research
project but CUEA failed to deliver an applied
science product it had promised regarding
management of living marine resources. What
didnt CUEA deliver? Why not? Is progress being
made?
5
According to Barber Conclusions regarding
useful to mankind forecasting 1. Goal was
inherently unattainable in 1972 to 1980. 2.
Limitations (deficiencies) in both in theory and
technology. 3. The deficiencies in theory (food
web structure, Fe, remote forcing, decadal
variability) were serious, but a lot (75?) of
the 1972/1980 physical and biological theory was
correct. 4. Technological limitations in
computation, observing systems and information
handling infrastructure were fatal due to the
constraints they imposed (undersampling,
inadequate space and time resolution and
inadequate model complexity.) 5. These technical
limitations were unconceivable at the time and
had to change many orders of magnitude before
useful forecasting could be done.
But the goal is now within reach
6
Science at the leading and/or bleeding edge
First ever long term forecast of chlorophyll?
7
Basin-scale model run for 10 years forced by NOAA
blended winds then forced by NCEP 9 month
forecast
Comparison of forecast anomalies with SeaWIFS
anomalies
We know forecast drifts after 5 months
Barber, Chai, Chao, Chavez
8
Approach

• Retrospective analysis (of in situ and remote
  sensing data)
• Identify changes in ecosystem and environment
• Develop conceptual and numerical models
• Look at model solutions (physics,
  biogeochemistry, fish) for drivers
• Forecast and hindcast

9
It is a familiar story
Once ever 3-8 years
Child
El Niño
La Niña
El Viejo
La Vieja
Parent
El Viejo
La Vieja
Once ever 25-40 years
1900 to 2000
10
Change
Two Primary States
Varia- bility
Low oxygen
11
Computing power allows for model simulations at
the right scale
Yi Chao, JPL
Realistic model of the Pacific at 12.5 km
resolution - SST
12
(No Transcript)
13
Yamanaka et al. (2005)
14
(No Transcript)
15
ROMS-CoSINE (12 km) Temperature,
Currents, Plankton
ROMS-CoSINE (12 km) Temperature,
Currents, Plankton
Life Cycle of Peruvian Anchovy Individual Based
Model with ROMS-CoSINE
ROMS-CoSINE (12 km) Temperature,
Currents, Plankton
ROMS-CoSINE (12 km) Temperature,
Currents, Plankton
16
Need to gain better access to US ocean ecosystem
scientists and their data and if data not
available then determine how to collect
17
Case Studies

• The Right Whale
• The Sablefish in Gulf of Alaska
• Leatherback turtles
• Productivity in the Arabian Sea

18
Right Whale, Wrong Time?

• Only 350-400 right whales in N. Atlantic
• Recovery is limited by hight mortality
• Ship strikes
• Fishing gear
• All management options depend on knowing where
  whales are

Andy Pershing et al.
18
19
Saving the Whales

• During spring, summer, and fall, whales follow
  food
• How can we find whale food on operational time
  scales?
• Use remote sensing of SST and Chl coupled with
  model circulation to predict Calanus

Andy Pershing et al.
19
20
Predicting Whales from Calanus
Whale Arrival Date
Whales arrive early when food is abundant
Andy Pershing et al.
20
21
Alaskan Sablefish ProjectAuthors S. Kalei
Shotwell Dana H. Hanselman

• Sablefish (Anoplopoma fimbria)
• Fast growing, wide distribution, highly valuable
  commercial species
• Adults generally at 200 meters in continental
  slope, gullies, fjords
• Early life history (ELH) largely unknown
• Spawning at depth (400m), larvae swim to
  surface, collect at shelf break
• Juveniles move nearshore to overwinter, then
  offshore in summer
• Reach adult habitat and recruit to fishery or
  survey in 4 to 5 years
• Recruitment calculated in age-structured model
• Recruitments are estimated as two year-olds
• Estimates for most recent years are highly
  variable with large uncertainty and excluded from
  model projections
• Objective
• Evaluate ELH data and explore integrating
  satellite derived environmental time series into
  the sablefish stock assessment to reduce
  recruitment uncertainty

22
Distribution, Movements, and Behaviors of
Critically Endangered Eastern Pacific Leatherback
Turtles Conservation Implications for an
Imperiled Population
Shillinger, Palacios, Bailey, Bograd Block Lab,
Stanford University NOAA-SWFSC-ERD
23
Chlorophyll (mg m-3) 10 year Sea-WiFS
Mean Kinetic Energy (cm2 s-2)
Surface velocities cm s-1
(CHL vs. Speed linear regression ? 0.964
0.057, F1,9577 281, P lt 0.001, r2 0.029)
CRD (10cm s-1)
NECC(30cm s-1)
SEC (n) (30cm s-1)
EUC (5cm s-1)
SEC (s) (15cm s-1)

• Turtles move into zones of low phytoplankton
  density

• Turtles must negotiate gauntlet of zonal currents

24
INTERANNUAL TRENDS IN PHYTOPLANKTON DYNAMICS IN
THE ARABIAN SEA LINKED TO EURASIAN WARMING
R. M. Dwivedi Space Applications Centre, Indian
Space Research Organization, India
25
SW Eurasian-Land Warming
Warming of SW Eurasia mirrors the global-land
signal, but recent warming anomalies are gt50
larger than global temperature trends.
26
Interannual changes in chlorophyll along coast of
Somalia since 1997 (Goes et al., Science, 2005 )
27
High chlorophyll concentrations during the NEM
are being caused by blooms of dinoflagellate
Noctiluca miliaris (not all chlorophyll is the
same!!!!)
A similar phenomena in Monterey Bay where (as in
the Arabian Sea) the nutricline is shallow and
dinos vertically migrate to depth at night for
nutrients and surface during day to take up
carbon dioxide
28

• Monterey Bay Time Series
• - El Niños during 92-93 and 97-98
• Transition from El Viejo to La Vieja
• The age of dinoflagellates?

Dinoflagellate regime associated with failures in
fish and seabirds
29
Longer Centennial changes (in oxygen)
Export production
Oxygen at 150 m
Gutierrez et al. - Paleopeces
30
Summary, during LIA ocean off Peru high
oxygen/few fish, low oxygen/high fish after
31
The low oxygen expanded southward in to Chile,
what about the recent record (50 years)

• California
• Peru

Stramma, L., G.C. Johnson, J. Sprintall, and V.
Mohrholz (2008), Expanding oxygen minimum zones
in the tropical oceans, Science, in press. May 2
32
Long-Term Trends in Dissolved Oxygen off
California
-2.1 ?mol/kg/y
DZmean -41 m
DZmax -92 m
Expansion of Low-Oxygen Habitat
(Bograd et al. in press)
33
In situ oceanographic data from Peru
1000 (Abstract ID2647) MondayMessié, M
Calienes, R Ledesma, J Barber, R T Pennington,
J T Chavez, F P INTERANNUAL VARIABILITY AND
LONG TERM TRENDS IN EASTERN PACIFIC UPWELLING
ECOSYSTEMS
34
It appears as if the eastern Pacific low oxygen
regions reformed after the Little Ice Age and
continue to expand today

• Are there biological indicators of this expansion?

35
The Hake off Peru has retreated and gotten more
concentrated
Hake habitat restricted by oxygen
Merluza durante La Niña (1996)
Hake inEcuador
Index of hake concentración
36
(No Transcript)
37
Post 1997/98 expansion of Dosidicus gigas range
2004
2001
July 2005 Tracy Arm (Sitka), AK
1984
2004 Outer Coast, BC
La Jolla Cove, CA. July, 2002
British Columbia Sept. 2005
Long Beach, WA Oct 2004
38
Including variability and change in management

• Has not been the norm for management of exploited
  populations typically managed using population
  models that do not parameterize the environment.
• Needs to be built in early on in our new
  ecosystem based management approach

39
Conclusions

• Sophisticated observing systems, increased
  computing power for modeling and improved
  ecosystem theory provide the framework today to
  develop ecological forecasts useful for society
• Nature will continue to surprise (at a faster
  rate due to global change) and our theory will
  need to constantly evolve
• NASA will need to develop new direct or indirect
  estimates (and models) of ecologically relevant
  properties beyond chlorophyll
• Better working relationships between resource
  managers and NASA should be fostered

40
Where climate meets the global economy
41
Global Climate
Global economy
Small Pelagic Fish
6Mt Fishmeal 1.2 Mt Fish oil (from 30 Mt fish
or 25 Global catch)
Poultry
Aquaculture
Hogs