NASA Biodiversity and Ecological Forecasting Team Meeting

1
We are NASA GRT NNG04GM64G Pacific climate
variability and its impact on ecosystems and
fisheries a multi-scale modeling and data
assimilation approach for nowcasting and
forecasting.
My name is Dick Barber (rbarber_at_duke.edu)
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
Fig.2
2
Dick Barber Duke University
Fei
Chai - University of Maine

Yi Chao - Jet Propulsion Lab of Cal Tech

Francisco Chavez - Monterey Bay Aquarium
Research Institute Joaquim Goes -
Bigelow Laboratory for Ocean Sciences
Michael Alexander - NOAA Climate Diagnostics
Center
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
Fig. 3
3
Goal Deliver forecasts of circulation and
ecosystem function in the Pacific Ocean which can
force fish population models to make forecasts of
fish abundance with 6 to 9 months lead-time.

NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
Fig.4
4
30 years ago (1975) we had an NSF
project, Coastal Upwelling Ecosystems Analysis
(CUEA), based on the premise that Prediction
of the response of the coastal upwelling
ecosystem to natural variations, man-made
environmental perturbations or to different
harvesting strategies is possible from knowledge
of a few biological, physical and meteorological
variables...
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
5
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. Written in May 1975 Little
did we know
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
6

• CUEA was successful by NSFs basic research
  criteria, but the goal was never achieved ...
• the technology, vision and science of 30 years
  ago was simply inadequate
• - undersampling in time and space
• - omission of remote forcing in the CU paradigm
• complete unawareness of decadal variability
  - over-simplified linear food web theory
• - limited to 2D and uncoupled modeling

NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
7
1. Two conceptual advances in
ocean ecological theory 1a.
two-path food web (need for picophytoplankton and
micrograzers) 1b. role of Fe in ocean
ecosystems
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
8
What has changed ? The ecology of 1975 was
pretty good, only two conceptual advances in
ocean ecological theory 1a.
two-path food web (need for picophytoplankton and
micrograzers) 1b. role of Fe in ocean
ecosystems
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
9
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
10
But the big change is 2. A revolution in
observing systems in
mode, resolution quantity. 3. An even more
profound revolution in
computational power.
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
11
2. The observational revolution (satellites and
moorings) gives continuous access to the time
and space scales of variability forcing ocean
ecosystems.
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
12
3. The computational revolution makes possible
and enormous increases in both time and space
resolution (physical) and model complexity
(ecology).
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
13
Project Columbia and ROMS at NASA Ames
12.5-km
Computer at NASA Advanced Supercomputing
Facility 20 interconnected SGI Altix
512-processor systems a total of 10,240 Intel
Itanium 2 processors Pacific basin-scale ROMS
(1520x1088x30) 12.5-km horizontal resolutions
30 vertical layers 50-year (1950-2000)
integration
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
14
that is, NASA satellites and computational-power
enable us to observe the ocean
(earth) continuously, initialize and assimilate
the observations into models with the
time/space resolution and
complexity needed for accurate
ecosystem and fisheries forecasts.
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
15
Scale convergence of eddy kinetic energy of
model and observations in a coastal upwelling
system

• Internal/intrinsic variability
• Features (lt10 km, days)
• Model resolution (1 km, hours)

Eddy kinetic energy (cm2s-2)
Drifter
Model
Resolution (km)
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
16
Assimilation and Initialization
3-dimensional variational (3DVAR) method
y observation x model
J 0.5 (x-xf)T B-1 (x-xf) 0.5 (h x-y)T R-1 (h
x-y)
17
Now, for a look at the Peru coastal upwelling
ecosystem Time series 1955 2005 Anchovy
catch by the Peruvian fishery Temperature
anomaly (C) in Niño 1 and 2 regions (3-month
running mean of monthly SST anomalies)
Niño 2
Niño 1
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
18
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
19

• Natural interannual (ENSO) and decadal (PDO, NAO)
  climate variability, together with fishing
  itself, drive large variations in fish stocks.
• Inability to account for externally-driven
  variability in fish stocks prevents successful
  management.
• Tools are now available to incorporate climate
  effects into ecosystem-based management models.
• Lead time is needed for economic usefulness.
• Operational 6 to 9 month physical forecasts (for
  the Pacific) are now available.

20
Show movie.
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
21

• We have
• Improved food web theory (Fe and 2-path)
• Realistic and validated physical and food web
  models
• Observing tools, satellites, moorings, TOGA-TAO,
  etc. for initialization and assimilation
• 4. Computational power needed for scale
  convergence, fine time steps and many model
  compartments
• 5. Operational 6 and 9 month ENSO forecasts

NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC
22
Summary The capability exists to deliver
operational marine ecosystem forecasts with
enough lead time, accuracy and precision to be
socially and economically beneficial.
NASA Biodiversity and Ecological Forecasting Team
Meeting 29 31 August 2005 Washington, DC