INSIGHTS FROM LABORATORY AND AIRBORNE BRDF MEASUREMENTS FOR SATELLITE REMOTE SENSING

1
INSIGHTS FROM LABORATORY AND AIRBORNE BRDF
MEASUREMENTS FOR SATELLITE REMOTE SENSING
Charles K. Gatebe,1 Michael D. King2 Georgi T.
Georgiev,3 and James J. Butler2
1University of Maryland, Baltimore County 2NASA
Goddard Space Flight Center 3Science Systems and
Applications Inc. (USA)
2
Outline

• Introduction
• background
• Example Results
• CAR Measurements
• Laboratory measurements
• Comparisons (Lab, Airborne, and satellite)
• Conclusion
• Future
• Acknowledgments

3
What is BRDF of these Surfaces?
WATER CLOUD(Namibian Coast, 9/13/2000)
SAVANNA (Skukuza, South Africa, 6/22/2005)
SALT PAN(Etosha Pan, Namibia, Landsat, 9/11/1999)
OCEAN (Chesapeake Lighthouse, 20 kmfrom
Virginia coast, 7/14/2001)
4
Instrumentation Used
Airborne - CAR

• Sensor Characteristics
• 14 spectral bands 0.34 to 2.29 µm
• scan 95 from horizon on right-hand side of
  aircraft
• field of view 17.5 mrad (1)
• scan rate 1.67 Hz (100 rpm)
• data system 9 channels _at_ 16 bit
• 395 pixels in scan line
• Scatterometer Characteristics
• Light source (75W Xenon short-arc lamp or HeNe
  Laser)
• Chromex 250SM scanning monochromator
• Goniometer - Sample stage (movable in X, Y, and Z
  linear directions), light source, detector

Laboratory - Goniometer
Light source
detector
5
BRDFs from Airborne Measurements
6
Do these pictures look different?
1204 p.m. SZA22,SAZ133
907 a.m. SZA54, SAZ88
418 p.m. SZA42, SAZ262

• Are differences due to BRDF effects only?
• What is the nature of BRDF of an individual leaf?
• Whats the nature of BRDF of the background
  surface (soil, leaf litter, etc)?

7
Samples for Laboratory BRDF
Fresh leaf
Fresh leaf
Dry leaf
Soil leaf litter
8
Fresh and Dry Leaves BRDF
SZA 0 (o) 30()
500 nm
900 nm
green fresh leave red dry leave
BRDF (sr-1)
View Zenith Angle ()
View Zenith Angle ()

• Leaves are from White Oak (Quercus Alba)
• Shapes of BRDF from fresh and dry leaves look
  different

9
BRDF Leaves and Soil
Leaf litter (870 nm)
Soil 1 (870 nm)
BRDF (sr-1)
View Zenith Angle ()
View Zenith Angle ()

• Shape of BRDF is similar for both leaf litter and
  soil at 870 nm
• Higher backscattering peak for the longer
  illumination zenith angle
• Most canopy reflectance models ignore soil
  reflectance anisotropy and without hot spot

10
Comparing Lab, Airborne and Satellite Measurements
Etosha Pan

• CAR MODIS measurements are normalized to the
  TOA solar fluxes, Lab meas are normalized to
  incident power
• Need to investigate if observed differences are
  real or are due to differences in IFOV, surface
  heterogeneity, etc.

11
Comparing Lab, Airborne and Satellite Measurements

• Removal of the effects of atmospheric absorption
  and scattering from CAR measurements resulted in
  over correction at 470 nm (Comparison not
  included)
• Laboratory brdf include only leaf litter
• Soil samples had much lower values (reflectance
  5)
• Important to understand the processes controlling
  BRDF

Savanna vegetation
12
Conclusion

• Apparent differences in BRDF between fresh and
  dry leaves for different incident angles are
  interesting
• Systematic differences between laboratory,
  airborne and satellite measurements especially
  for salt pan (bright target) are noted
• Better agreement between airborne and laboratory
  especially for savanna vegetation (relatively
  darker target), ignoring soil reflectance

13
More Measurements
Future

• Long-term goal of this work is to try to
  understand BRDF from laboratory, airborne and
  satellite measurements and then attempt to unify
  them with a consistent model

Satellite - Terra
Laboratory - GSFC
Unified Model ???
Airborne - J-31/CAR
14
Future Measurements BRDF OF Complex Surfaces
Mexico City
1043 a.m. 5740 m SZA25
1104 a.m. 3448 m
1112 a.m. 2501 m
1112 a.m. 2501 m
15
Future
Simultaneous retrieval of surface BRDF and
aerosol properties from CAR will reduce
uncertainties both in aerosol and surface BRDF,
and will allow us to handle even complex surfaces
16
Acknowledgements
Aircraft Peter Hobbs UW CARG Piketh, Annegarn
teams Ben Hovelman, Sky Research
Programming Tom Arnold Gala Wind
Atmospheric Correction Alex Lyapustin Eric Vermote
Validation Steve Platnick Wedad
Abdou/MISR Phillip Russell AATS team
Calibration Jim Butler/John Cooper Scott
Janz/Matt. Kowalewski Jeff Myers et al.
Web Goran Harusa Paul Hubanks/Jason Li
Engineers Peter Shu/Hollis Jones et al.
Funding EOS Hal Maring/Don Anderson
Inversion Oleg Dubovik Alexander Sinyuk