ecological observation modeling approach
DESCRIPTION
Ecological Observation Modeling Approach. Dennis Ojima Natural Resource Ecology Laboratory. MAY 2006 Tucson, Arizona. Collaborators David Schimel (NCAR), Steve Running (U of MT), Russ Monson (CU), Brit Stevens (NCAR), Jeff Hicke (CSU). Funding from NSF, NASA, NOAA. WHY NOW?. - PowerPoint PPT PresentationTRANSCRIPT
Ecological Observation ModelingApproach
Dennis OjimaNatural Resource Ecology Laboratory
MAY 2006Tucson, Arizona
CollaboratorsDavid Schimel (NCAR), Steve Running
(U of MT), Russ Monson (CU), Brit Stevens (NCAR), Jeff Hicke (CSU)
Funding from NSF, NASA, NOAA
WHY NOW?WHY NOW?• Grand Challenges facing Environmental
Sciences– Land Use; Climate change; Biodiversity;
Biogeochemical cycles; Infectious disease; Invasive species
• New Observations for Terrestrial Systems• New Cyber Infrastructure Developments• New Advancements in Quantitative Analysis• Development of Data-Model Fusion
Techniques
Land Use Change
More land was converted to cropland in the 30 years after 1950 than in the 150 years between 1700 & 1850
Cultivated Systems in 2000 cover 25% of Earth’s terrestrial surface
(Defined as areas where at least 30% of the landscape is in croplands, shifting cultivation, confined livestock production, or
freshwater aquaculture)Millennium Ecosystem Assessment 2005
Temperature Anomalies (2003)
Data source: (Jones and Moberg 2003). Processed by the U.S. NOAA NCDC Global Climate at the Glance Mapping System
MODIS Optical Density (Average 2001)
Global and Regional Telecommunications
Fire
Dust Storms
Nitrogen
BrownCloud
NitrogenNitrogen
CHALLENGES and NEEDSCHALLENGES and NEEDSOF OF
Terrestrial Environmental Terrestrial Environmental Observations and AnalysisObservations and Analysis
• Multiple Stresses• Interactive Sectors• Increasing Human Pressures
• Information Exchange to Multiple Publics– Science– Managers– Policy Makers– Public at Large
This is a Time of Great Opportunity
• Digital information explodes
• Bandwidth increases
• Wireless capabilities expand
• HPCC and IT technologies advance and pervade science and society
• Collaborative, multidisciplinary activities increase
• Integrative approaches demanded
FROM PETABYTA TO SOUNDBYTE
Multi-sensor/Multi-scale Modeling Framework
Nemani et al., 2003, EOM White & Nemani, 2004, CJRS
• Collect data from digital libraries, laboratories, and observation
• Analyze the data with models run on the grid
• Visualize and share data over the Web
• Publish results in a digital library
Changing How Science is Done
NEON: A continental research platform designed to provide the capacity to forecast future states of ecological systems for the advancement of science and the benefit of society
Nat
ion
al E
colo
gic
al O
bse
rvat
ory
Net
wo
rk (
NE
ON
)
Novel infrastructure that:
• allows scientists to observe the previously unobservable
• scale from m2 to continent
• evaluate fundamental theory at regional to continental scale
• enables a new forecasting and predictive capacity for ecology
• takes advantage of new and evolving in situ sensing technologies
• couples human and natural systems
From points to pixels
?
Create high res. productsby coupling high res. imagerywith field and tower data
Aggregate
Correlate
Some graphics courtesy of BigFoot project, layout courtesy of Shunlin LiangMultiple use of airborne or high res. satellite data for extrapolation of sites observations
• SCIENCE BASED: Developing and testing theory and models requires integration of complex in situ process data with large gridded data sets.
• MULTI-SCALED: Required data are multi-scale, many formats, originating in multiple disciplines.
• AGILE: Rapid prototyping and development cycle to maximize user control of information systems, implies incorporating existing state-of-the-art components rather than de novo development
• USER-DRIVEN: Data systems must allow user-driven, knowledge-based querying of multiple data types
Information Technology for Biogeosciences
ACME-CME: (Aircraft) C in the
Mountains Experiment
Sponsors: NSF-NASA Collaborating Inst: CU-NCAR-CSU-NOAA
ACME-CMEACME-CME• To understand carbon dynamics in
montane forest regions by developing new methods for estimating carbon exchange at local to regional scales
• “Bottom-Up” (plot and tower obs) and “Top-Down” (aircraft and satellite obs) constraints
• Evaluate factors affecting C-exchange in complex terrain as compared to flat landscapes to better understand the significance and contribution of mountain areas to the continental carbon budget.
IntegratingAcrossScalesThroughTop-down&Bottom-upApproaches
F
VOCCO
VOCCO
CO2
Biogenic sourceMissoula: Urban source
CH4
Carbon-containing pollutant transport: 10s-1000s of km
Wildfire source
Downslope flows and subsequent venting of CO2
Tower
Soil Chamber
WindWind
Upw
ind Profiles
Erosion and organic matter transport
Dow
nwind P
rofiles
NE
EFootprint
Courtesy of Steve Running, U Montana
Model-data fusion: processes at the scale of biosphere-atmosphere
exchange
PLANT WOOD CARBON
PLANT LEAF CARBON
Photosynthesis Autotrophic Respiration
Leaf Creation
VEGETATION
SOIL CARBON
Wood Litter Leaf Litter
Heterotrophic Respiration
Precipitation
Niwot Ridge, Colorado
SipNET
Airborne Carbon in the Mountains Experiment
SPACENET Model Structure
Plant Carbon
Soil CarbonSoil Moisture
Drainage
Precip. Transpiration
Photosynthesis (Phenology,Soil Moisture,
Tair, VPD, PAR)
Plant Respiration(Plant C, Tair)
Litterfall(Plant C, Phenology)
Soil Respiration(Soil C, Soil Moisture,
Tsoil)
NE
E (
g C
/m2 p
er h
alf-
dail
y ti
me
step
)
NE
E (
g C
/m2 p
er h
alf-
dail
y ti
me
step
)
Initial guess parameters Optimized parameters
Blue: ModelRed: Data
Dates: 11/1/98 – 10/31/02Each point represents one half-daily time step
Model vs. Data: Unoptimized & Optimized Parameters
Cum
ulat
ive
NE
E (
g C
/m2 )
Cum
ulat
ive
NE
E (
g C
/m2 )
DODS Aggregation
Server
GrADS-DODSServer
Regional carbon data assimilation system based on RAMS atmosphere model
4D VARAssimilation
System
Compare, minimize
Estimated Fluxes
ObservingSystem
CO2
Observations
http-BasedInterface
4D VAR Assimilation
System
RAMS
Optimizer
RAMS Adjoint
CO2,
And Meteorological Observations
1st Guess fluxesFrom
SPACENET
Airborne and surface CO2,
obs
Carbon Data-Model Assimilation (C-DAS)
http://dataportal.ucar.edu/CDAS/
Multiple Modeling Approach Multiple Modeling Approach for Terrestrial C Fluxesfor Terrestrial C Fluxes
1) Inverse modeling of atmospheric chemistry•Constrain C sinks; little info on why or exactly
where2) Biogeochemical models
•Test physiological changes; little land use or disturbance
3) Land-use bookkeeping models•Track land-use change, but no climate or
physiology4) Flux towers
• Integrated site-level measurements, but relatively few sites
5) Forest inventories
ACME/CDAS ApproachACME/CDAS Approach
• Bring together observationalists and modelers to form an integrated approach to improving our understanding of the global carbon cycle.
• Initial effort: Network design exercises based on a selected assimilation modeling strategy.
• Ongoing: Further development of the assimilation tool and support for testing and planning/educational use by the community.
• Developing and testing theory and models requires integration of complex in situ process data with large gridded data sets.
• Required data are multi-scale, many formats, originating in multiple disciplines.
• Rapid prototyping and development cycle to maximize user control of information systems, implies incorporating existing state-of-the-art components rather than de novo development
• Data systems must allow user-driven, knowledge-based querying of multiple data types
Information Technology for Global Environmental Change Sciences