Integration Across Scales: The Role of Remote Sensing

1
Integration Across Scales The Role of Remote
Sensing Models
Warren B. Cohen, PNW Research Station USDA,
Forest Service
Scalable Information Networks for the Environment
(SINE)29 to 31 October 2001 San Diego
Supercomputer Center
2
Biosphere Monitoring

• Study Earth system at various scales (grains
  extents)
• Toolset includes field data, image-based remote
  sensing, models, conceptual framework for
  integration
• Status of trends in remote sensing
• Examples of integration

3
Scales scaling review of tools concepts

• Emergent properties are focal point for entry
  into scaling studies
• We seek to understand processes that help
  explain the
• existence or state of observed emergent
  properties of systems under study

4

• Remote sensing a scaling tool for looking at
  the Earth from above (inverse macroscope)

• Models/modeling scaling tools to understand
  complex interacting processes within the Earth
  system

5
Remote Sensing status trends in the discipline
6
Passive Optical Sensors
7
Trends in technology hyperspectral data
8
Trends in technology hyperspatial data
9
Trends In Passive Optical Imaging Sensing
Hyper- spectral
Global mapping
Hyper- temporal
Regional mapping
Hyper- spatial
Local mapping
10
New kid on the block waveform sampling lidar
(active optical sensing)
11
(No Transcript)
12
Temperate coniferous
Boreal coniferous
Temperate deciduous
1400
Temperate Deciduous
1200
Temperate Coniferous
Boreal Coniferous
1000
800
Observed Biomass (Mg ha-1)
600
400
Non-asymptotic predictions of biomass, LAI, etc.
200
0
-200
-200
0
200
400
600
800
1000
1200
Predicted Biomass (Mg ha-1)
13
Lidar

• Canopy volume
• Canopy organization
• Canopy surface morphology

14
Remote sensing models a generalized framework
for scaling in ecology
Site Measurements
a, b
b
Spatial Data Algorithms
Ecological Models
b
Validation
b
Scaled Ecological Information
a
a, b
15
a) Land cover mapping
Accuracy 80
Cover () b0 b1 greenness
b2 brightness Conifer () b0 b1
brightness b2 wetness
Age (yrs) b0 b1 wetness
b2 greenness
16
a) Stand replacement disturbance mapping
Accuracy 92
17
b) C-flux modeling
18
Global Biosphere Monitoring a first step
Highly generalized data, algorithms, models
first approximation useful for specific purposes
19
another important component ground data
20
Conceptual framework for extrapolation
21
Current configuration of BigFoot sites
22
Sampling design, nested from within plot to
greater tower footprint, across sites
23
(No Transcript)
24
(No Transcript)
25
MODIS NPP ???
26

• Crop type
• matters
• Inter-annual
• variability
• important

27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
Global Biosphere Monitoring
Where does a NEON-like network fit?
NEON-like network
33
Summary

• Scaling field-based ecological information in
  space and time requires an integrated strategy,
  geographic databases (e.g., remote sensing,
  climate), spatially- temporally-explicit
  models, and a conceptual framework
• Remote sensing is undergoing rapid advancement,
  with convergence toward increasing spatial
  spectral resolutions, and better quality high
  temporal resolution data
• NEON-like network can play a major role in
  global biosphere monitoring, especially given the
  broad range of spatial scales such a program will
  represent, the length of time the program will
  likely exist, and the importance of detailed
  processes it will study