3. Remote sensing of forest vegetation (via MODIS enhanced vegetation index, EVI)

shows “green-up” during 2005 drought, opposite to model-predicted response in “hypothesis” above

U.S. University of Arizona: Scott Saleska (Ecol. & Evol.

Biol, EEB.), James Shuttleworth (Hydrology), Alfredo Huete (Soil, Water & Env. Sci.); Judie Bronstein (EEB); Brian Enquist (EEB); Tim Finan (Anthropology); Travis Huxman (EEB); Joellen Russell (Geosciences); Sylvia Tesh (Latin Am. Studies); Maria Teresa Velez (Grad. College); Scott Whiteford (Latin Am. Studies)

BrazilHarvard University: Steven Wofsy (Earth & Planet. Sci)Paul Moorcroft (Organismal & Evol. Bio.); Scot Martin (School of Engineering & Appl. Sciences

University of São Paulo: Plinio de Camargo (CENA); Paulo Artaxo (Inst. of

USFS / NEON: Michael Keller

Physics); Humberto da Rocha and Pedro L. Silva-Dias (Atm. Sciences); Federal University of Pará: Julia Cohen, Renato da Silva, and Rodrigo da Silva (Meteorology); National Institute for Amazon Research (INPA): Antonio ManziBrazil Agriculture Research Agency (EMBRAPA): Raimundo de Oliviera Museu Goeldi and National Inst. for Space Research (INPE): Leonardo Sá

for International Research… … and Education:Question:

What is the future of Amazon forests under climate change?

Hypothesis:

Hypothesis: this feedback implies water-limitation: ET and photo-synthesis should be reduced during dry periods, especially strong drought.

This hypothesis is testable under present climates by observing forest response (fluxes, vegetation dynamics, and remotely sensed indices) to real-world ENSO-like drought, and comparing to (B) model-predicted responses to individual ENSOs at local study sites (Jones et al., 2001).

Model Tropical Forest

(A) spatially distributed canopy structure observations (e.g. remote-sensing Lidar) are used to constrain:

(B) size and age-structured model (Ecosystem Demography model, ED), which

(C)makes carbon balance predictions testable against

(D) temporally and spatially distributed observations at the tower and across the surrounding National Forest.

6. Constrain & test models linking ecosystem structure and biogeochemistry at multiple scales (Models constrained by

remote-sensing make testable predictions across time and space)

1. Forest metabolism (via eddy covariance measurements at a network of tower sites)

Aircraft-based LIDAR: Aircraft LIDAR surveys planned for this project will retrieve canopy structure over large spatial scales, and can constrain structure-based ecosystem models at those large scales (e.g. as in Hurtt et al., 2004)

Drought Experiment in Model Tropical Forest

Training and education

Modeling andhypothesis generation

Observationsin Amazon forests

Team building, methods development for programmatic success

Idea generation, hypothesis testing, for knowledge building

Developing tools and personnel

Developing methods, ideas and models

Goal: International and Interdisciplinary Earth & Ecosystem Science Education

Conceptual model for international research and education:

Concepts and Tools for simultaneously developing knowledge and international teams

Concepts

Tools Teams

Knowledge

1.Interdisciplinary curriculum & advising:

Scientific coursework encompassing (a) ecology & biogeochemistry; (b) hydro- meteorology & climate; (c) ecosystem modeling

Instruction in Portuguese language and Brazilian culture, including core interdisciplinary PIRE course, “Science, Politics, and Culture of Amazon Landscapes”

Effective mentoring by U.S. and Brazilian advisers and collaborators

2.Amazon field course and Biosphere 2 model tropical forest: foci for development of international student research projects Project-based field course as pipeline to ongoing field research and collaboration with Brazilian partners. Students collaborate with both world-class Brazilian scientists and local Amazon woodsman (mateiros) whose main education comes from working in the forest.

3.International collaboration via Exchanges with Brazilian field sites, institutions and laboratories U.S. students will work at Amazon field sites, including the Tapajós National Forest (near Santarém), the Caxiuana National forest (near Belém), and Cuieiras Reserve (near Manaus). World-class laboratories (e.g. for isotope analysis) and computational facilities for modeling and remote sensing. Brazilian students will also participate, through exchange (via funding from Brazil or other sources) with U.S. Institutions.

4.Broader Education & OutreachLocal education/outreach presentations in Amazonian universities by PIRE participants will enhance community relations in Amazônia, and provide a basis for educational outreach in the U.S. about global environmental issues.

Four-part International Education program:

• The focus is on graduate education, but advanced undergraduate projects/senior theses will also be considered for support.• Amazon-PIRE fellowships include a competitive annual stipend (30K per year, for one to two years), full tuition, health insurance, and allowance for travel to Brazilian field sites and collaborating institutions.• Amazon-PIRE is committed to diversity in education, and particularly encourages the application of women and underrepresented minorities.• Eligibility: U.S. citizens or permanent residents currently enrolled in earth system science-related graduate studies at University of Arizona or Harvard University.

4. Aircraft campaigns(Aircraft observations of forest structure at landscape scales)

5. Experiments in Biosphere 2 (Large-scale Drought manipulations in model tropical forest of Biosphere 2)

Contact: www.eebweb.arizona.edu/pire_amazoniaScott Saleska (saleska@email.arizona.edu)

Maria Theresa Tostes-Ribeiro McLaughlin (mttrm@email.arizona.edu)

ReferencesBetts et al., (2004). Theor. Appl. Climatology, 78, 157-175.Cox P.M., et al. (2000). Nature, 408 (6809): 184-187.Dufresne (2002). Geophys.l Res. Lett., 29 (10), 1045.Hurtt et al., (2004). Ecol. Appl. 14 (3): 873-883.Jones et al. (2001). J. of Climate, 14 : 4113-4129.Saleska et al. (2007). Science, 318: 612.

Amazon-PIRE: a Partnership …

Field Course 2008:July 12-23, 2008 at Tapajós National Forest

(near Santarém, PA, Brazil)

pre

cip

itat

ion

ano

mal

y

veg

etat

ion

“g

reen

nes

s” a

nom

aly

2. Vegetation dynamics (via biometric surveys on spatially extensive permanent plots)

Mateiro assisting in tree survey, in 20 ha plot in the footprint of eddy tower near Santarém (3000 trees + dead wood)

Units: number of standard deviations in 2005 from the long-term mean for the July/Aug/Sept (JAS) quarter. I.e., for each pixel:

Sal

esk

a, D

idan

,H

uete

, R

ocha

(2

007)

The University of Arizona Biosphere 2 campus, in Oracle, AZ, has begun a new phase of integrated research and teaching for ecosystem science: http://www.b2science.org

Goal: Observe forest response to severe ENSO droughtTest Hypothesis by Integrating…

(A) Coupled carbon cycle/climate models of global terrestrial carbon sink

0

12

4

8

-4Emis

sion

s (

Gt

C yr

-1)

Terrestrialsink

Residual emissions (Fos.Fuel minusocean sink) Hadley

Center

I PSL

Loss of forest

Year: 1950 2000 2050 2100

0

12

4

8

-4

0

12

4

8

-4Emis

sion

s (

Gt

C yr

-1)

Terrestrialsink

Residual emissions (Fos.Fuel minusocean sink) Hadley

Center

I PSL

Loss of forest

Year: 1950 2000 2050 2100

(B) Hadley predicted C-flux & precip in central Amazonia through years relative

to El Nino drought

Drought

Fore

st e

fflux

(M

gCha

-1yr

-1)

Year: -2 0 2 4

0

10

0 2

00 (

mm

mo

-1)

Prec

ip

DroughtDrought

Fore

st e

fflux

(M

gCha

-1yr

-1)

Year: -2 0 2 4

0

10

0

10

0 2

00 (

mm

mo

-1)

Prec

ip

Fig. 1. (A) Models predict starkly different futures for the terrestrial carbon sink: from sink enhancement (IPSL, duFresne et al., 2002, in green) to switchover to C-source (Hadley Center, Cox et al., 2000, in orange), a consequence in part of catastrophic collapse of the Amazon forest, triggered by onset of a semi-permanent ENSO drought. The Hadley-predicted collapse requires amplification of drought effects by forest physiological feedback (Betts et al, 2004): initial drought+high CO2 stomatal closure reduced evapotranspiration (ET) reduced recycled precip more drought, etc.

Caxiuana

São Gabriel Cachoeira

Sinop

Equator

10 S

20 S

Amazônia

(A) Atlantic Ocean

Caxiuana

São Gabriel Cachoeira

Sinop

Equator

10 S

20 S

Amazônia

(A) Atlantic Ocean

Caxiuana

São Gabriel Cachoeira

Sinop

Equator

10 S

20 S

AmazôniaCaxiuana

São Gabriel Cachoeira

Sinop

Equator

10 S

20 S

Amazônia

(A) Atlantic Ocean

2005,2005,

JAS JASJAS

JAS

x xAnomaly

Eddy fluxes of CO2 and energy cross a network of towers ()

Eddy flux tower at km67 site in the Tapajós National Forest (showing PI on tower for scale)

NSF stipends to support U.S. student participation

in Amazon-PIRE:Including both full-year stipend support,

and stipends for summer projects.

Amazon-PIRE: What is the future of Amazon forests under climate change?funded by the National Science Foundation

PI: Scott Saleska, University of Arizona, Biosphere 2