GlobColour international context (as

1
GlobColour international context (as seen from
IOCCG) Prepared by IOCCG Project
Office Presented by Eric Thouvenot (CNES
representative to IOCCG)
2
IOCCGs perpective of International Context of
GlobCOLOUR
As operational oceanography grows, there is an
increased demand for data and information
relevant to understanding the marine ecosystem at
the global level. Many issues could be addressed
using the GlobCOLOUR data set, both global and
regional. Example 1 Yoder research -
long-term time series are required to examine
changes in global chlorophyll levels and to sort
out differences between cycles and trends.
Example 2 Platt research - long-term ocean
colour time series can explain haddock
recruitment fluctuations in the Northwest
Atlantic. There is great potential that the
GlobCOLOUR dataset will provide similar answers
elsewhere. The GlobCOLOUR dataset is also
relevant to several of the tasks of the
intergovernmental Group on Earth Observations
(GEO), which is leading a worldwide effort to
build a Global Earth Observation System of
Systems.
3
Example 1 Yoder slides
4
Are Phytoplankton Biomass and Productivity
Declining in Large Parts of the Global Ocean
Owing to Climate Change Effects on
Stratification? Jim Yoder Woods Hole
Oceanographic Institution
5
Questions

• Are model and SeaWiFS phytoplankton chlorophyll
  (Chl) trends similar during the SeaWiFS-era
  (1998-2005) in regions of the open ocean, i.e.
  does the model agree with the satellite
  observations?
• Are model trends unusual during the SeaWiFS-era
  compared to other 8-year intervals during the
  model era (1958-2004)?

6
Motivation and Background

• Three recent manuscripts indicate that
  phytoplankton chlorophyll/carbon (Chl)
  concentrations in large regions of the ocean are
  decreasing, possibly owing to climate change
  effects on ocean stratification.
• Gregg, W. et al. 2005, Geophys. Res. Lett., 32,
  L03606, doi 10.1029/2004GL021808.
• Antoine et al. 2005, J. Geophys. Res., 110,
  C06009, doi 10.1029/2004JC002620.
• Behrenfeld et al. 2006, Nature 444
  doe10.1038/nature0517

7
Trend in SeaWiFS Mean Chlorophyll (1997-2003)
(from Gregg et al. 2005)
8
Percent differences (red is 100, blue is
lt50) in SeaWiFS (1998-2002) to CZCS (1979-1983)
annual mean chlorophyll (from Antoine et al. 2005)
9
Trend in SeaWiFS chlorophyll, net primary
production (NPP) and stratification anomalies
(MEI) for stratified waters of the global ocean
(from Behrenfeld et al. 2006)
Chl (line)
MEI (circles)
Chl trend(line)
10
Impact of 96-97 ENSO on Satellite SST and Chl
Anomalies (from Yoder, J.A. and M.A. Kennelly.
2003. Global Biogeochemical Cycles, 17 (4), 1112.
). Note that SST and (inverse) Chl anomalies
track ENSO index.
low
Global anomalies (mean seasonal trend removed
from each pixel) summed from 50S to 50N.
Blue is negative Chl (i.e. 0 - Chl) Black is
SST. Red is Nino 3.4 index.
11
Approach

• Use the model to see if the trends observed in
  SeaWiFS imagery are also evident in the longer
  record (46 years) of the model.
• Answer the following questions
• Are model and SeaWiFS phytoplankton chlorophyll
  (Chl) trends similar during the SeaWiFS-era
  (1998-2005) in regions of the open ocean, i.e.
  does the model agree with the satellite
  observations?
• Do model results indicate that the SeaWiFS-era
  trends are representative of longer period trends
  showing decreasing Chl concentrations possibly
  linked to increasing ocean stratification?

12
Model References

1. Moore, J.K., S.C. Doney and K. Lindsay, 2004
   Upper ocean ecosystem dynamics and iron cycling
   in a global 3-D model, Global Biogeochem. Cycles,
   18, 4, GB4028, 10.1029/2004GB002220.
2. Doney, S.C., K. Lindsay, I. Fung and J. John,
   2006 Natural variability in a stable 1000 year
   coupled climate-carbon cycle simulation, J.
   Climate, 19(13), 3033-3054.
3. Doney, S.C., S. Yeager, G. Danabasoglu, W.G.
   Large, and J.C. McWilliams, Mechanisms governing
   interannual variability of upper ocean
   temperature in a global hindcast simulation, J.
   Phys. Oceanogr., in press.

13
Brief Model Description
Starts with the 3-D physics of the Parallel Ocean
Program (POP) with 3.6 long and 0.9 to 2.0 lat
resolution, and we used monthly fields. Forced
by NCEP reanalysis fields. Embeds a
multi-element, multi-functional group ecosystem
model (Moore et al. 2004) and a marine
biogeochemistry model (Doney et al.
2006). Primary production is partitioned between
pico and nano phytoplankton and includes diatoms
and diazotrophs. N-fixation and calcification
are calculated. Zooplankton include micro and
larger forms. Nutrients include N, P, Si and Fe.
14
Analysis Sites
11 (green dots) of our 34 stations are within
areas identified by Gregg et al as showing
significant trends in SeaWiFS Chlorophyll. Used
these sites to examine model and SeaWifS Chl
trends.
15
Conclusions

• Are model and SeaWiFS phytoplankton chlorophyll
  (Chl) trends similar during the SeaWiFS-era
  (1998-2005) in selected regions of the open
  ocean, i.e. does the model agree with the
  satellite observations?
• Yes.
• Do model results indicate that the SeaWiFS-era
  trends are representative of longer period trends
  showing decreasing Chl concentrations possibly
  linked to increasing ocean stratification?
• No, the model shows that there are 8-year
  periods of both increasing and decreasing
  chlorophyll trends throughout the model era
  (1958-2004).

16
Speculation

• Trends observed during SeaWiFS era are related to
  the very large ENSO which began in 1996 and had
  global impacts for many years (e.g. see Yoder,
  J.A. and M.A. Kennelly. 2003. Global
  Biogeochemical Cycles, 17 (4), 1112. ). Thus, he
  apparent changes (both increases and declines
  depending on ocean region may not reflect a long
  term trend.

Yoder and Kennelly 2003
Behrenfeld et al. 2006
17
Example 2 Platt slides
18
Example 2 Using ocean-colour remote sensing as a
tool for development of ecological indicators in
the coastal zone (Platt 2003, plus unpublished
data)
Ecological indicators are an aid to
ecosystem-based management essential to have
long time series. With ocean colour data,
construction of time series is possible at any
chosen scale of spatial averaging.
Time series in NW Atlantic note differences in
the timing of the spring bloom from year to year
and region to region. (SeaWiFS data)
19
Quantifying Seasonality
Any or all of these indices may vary between
years (at any or all of the pixels in the region
of interest)
Platt, Sathyendranath Fuentes-Yaco, 2007
20
Using time-series data to test Cushings
Match-Mismatch Hypothesis Test whether
significant proportion of variance in larval
abundance (survival) can be accounted for by
variations in ecosystem indices (interannual
fluctuations in dynamics of spring bloom).
Anomalies for Timing of Chlorophyll a Maxima
(February - July)
1997
1998
1999
2000
2001
Platt et al. 2003
21
Using time-series data from CZCS and SeaWiFS, it
appears that greatest larval survival coincides
with an earlier spring bloom.
Haddock survival in the NW Atlantic (after Platt
et al. 2003)
22

• Where number of haddock larvae and biomass of
  phytoplankton overlap, larvae have food supply
  adequate for survival
• Where this is not so, larvae are vulnerable to
  death by starvation

Early blooms imply a smaller blue area and a
smaller proportion of the total larvae produced
at risk from inadequate food supply
23
Conclusion

• Remotely-sensed data are useful for
  construction of time series, but requires care in
  quality control.
• Time series provide cost-effective basis for
  development of
• ecological indicators, averaged at appropriate
  time and space scales.
• Even with only two remotely-sensed variables
  (chlorophyll and temperature), a rich set of
  ecological indicators can be derived.
• The SeaWiFS 10-year series has yielded
  interesting results. The GlobCOLOUR data set will
  likely add more.

24
Part 3 GlobColour and GEO
25
GlobCOLOUR and GEO
Other Relevant International Projects

• The GlobCOLOUR dataset may also be relevant to
  the following GEO tasks
• ChloroGIN Project (Chlorophyll Global
  Integrated Network). GEO Task EC-06-07.
• The SAFARI Project (Societal Applications in
  Fisheries and Aquaculture using
• Remote Sensing Imagery). GEO Task AG-06-02.
• Global Ecosystems Classification and Mapping
  Initiative (GEO Task EC-06-02)

26
ChloroGIN Project (Task EC-06-07)
Chlorophyll Global Integrated Network

• Goals
• To develop a global network which will provide
  information on marine ecosystems for use at
  national and regional scales, using a combination
  of Earth observation (EO) data from satellites
  (Chl and SST) and in situ observations.
• To integrate in situ and remote observations
  into a single network - this will improve
  understanding of ecosystem processes and dynamics
  and will help in fisheries management.
• To provide a timely delivery of data and
  information that will benefit society.

27
ChloroGIN Project contd.

• Data delivery may be in near-real time (NRT) or
  delayed mode time series data (similar to
  GlobCOLOUR dataset, but not merged data).
• Latin American network (ANTARES) already
  established with the aim of studying long-term
  changes in coastal ecosystems. In situ and
  satellite data from around South America shared.
• ChloroGIN Africa web portal was recently
  established along the same lines as ANTARES.
• GlobCOLOUR dataset may help fill some of the
  gaps of ChloroGIN

ChloroGIN Africa
ANTARES (South America)
28
SAFARI Project (Task AG-06-02)
Societal Applications in Fisheries and
Aquaculture using Remotely-sensed Imagery
Project initiated October 2007 Funded by The
Canadian Space Agency Chairman Dr. Trevor Platt

• Project Goals
• To coordinate, at the international scale,
  various earth-observation initiatives related to
  fisheries and aquaculture, and add to their value
  through synergy.
• Project Execution
• Host an international coordination workshop
• Publish an IOCCG monograph on the state of the
  art
• Highlight excellent demonstration projects of
  EO in fisheries
• Develop an outreach component to increase
  awareness of the value of EO in the fisheries and
  aquaculture sector
• Convene an international symposium on this
  timely topic

29
SAFARI Demonstration Projects

• Examples of some elements that may be included in
  SAFARI and are relevant to GlobCOLOUR
• An internationally-coordinated programme in the
  Northwest Atlantic Ocean (SHRIMP) to relate
  relative abundance and growth of the Northern
  Shrimp to ecosystem fluctuations, as indexed by
  remote sensing of ocean colour
• A Canadian programme on development and testing
  of ecological indicators for the pelagic zone, as
  deduced from EO data, and evaluation of their
  utility for ecosystem-based management.
• The design of an ocean-colour constellation of
  satellites for long-term, uninterrupted,
  internally-consistent stream of remotely-sensed
  data for operational applications (initiative of
  the IOCCG).
• Southern African work on integrated,
  ecosystem-based and cooperative management of the
  Benguela ecosystem.

30
Global Ecosystems Classification and Mapping
Initiative (Task EC-06-02)

• Aim
• To establish an ecosystem classification task
  force, covering freshwater, terrestrial and ocean
  ecosystems, with a mandate to create a globally
  agreed, robust and viable classification scheme
  for ecosystems.

• In parallel with the classification effort,
  develop, review, and initiate a mapping approach
  to spatially delineate the classified ecosystems.

31
Global Ecosystems Classification and Mapping

• Global ecosystems can be classified at the
  meso-scale (on the order of 10 to 103 km2)
• A biophysical stratification approach can be
  adopted for terrestrial, freshwater and marine
  ecosystem delineation.
• For the oceans, the approach of Longhursts
  (1998) biogeochemical provinces can adopted.

Longhurst Partition
Boundaries may move seasonally (Devred 2007)
32
(No Transcript)
33
Summary

• We still lack sufficiently long satellite time
  series to sort out differences between cycles and
  trends.
• We need a sustained international effort to make
  sure we can link one satellite data set to
  another to build the long time series that we
  need.
• GlobColour is definitely a significant step in
  that perspective