A new remote sensing dataset for the lowveld M.M. Verstraete 1 and - - PowerPoint PPT Presentation

a new remote sensing dataset for the lowveld
SMART_READER_LITE
LIVE PREVIEW

A new remote sensing dataset for the lowveld M.M. Verstraete 1 and - - PowerPoint PPT Presentation

A new remote sensing dataset for the lowveld M.M. Verstraete 1 and R.J. Scholes 2 1 JRC-Institute for Environment and Sustainability 2 CSIR-Natural Resources and Environment The MISR instrument Earth surface viewed at 9 angles Four spectral


slide-1
SLIDE 1

A new remote sensing dataset for the lowveld

M.M. Verstraete1 and R.J. Scholes2

1 JRC-Institute for Environment and Sustainability 2 CSIR-Natural Resources and Environment

slide-2
SLIDE 2

Earth surface viewed at 9 angles

70.5º 70.5º 60.0º 45.6º 26.1º 0.0º 60.0º 45.6º 26.1º Forward-viewing cameras Forward-viewing cameras Backward-viewing cameras Backward-viewing cameras Df Da Ca Ba Aa An Cf Bf Af

The MISR instrument

Four spectral bands at each angle: 446 nm ± 21 nm 558 nm ± 15 nm 672 nm ± 11 nm 866 nm ± 20 nm

<7 minutes to view each scene from all 9 angles

Nadir camera Nadir camera Flight direction ~7 km/sec

slide-3
SLIDE 3

What is this thing called BRDF?

Bidirectional Reflectance Distribution Function

  • Solar radiation is a highly directional source of light
  • Objects scatter incoming light differently in different directions, due to

their intrinsic optical properties and overall geometric structure

  • BRDF models describe the anisotropy of the reflectance, given the

illumination and observation directions

  • Imaging instruments observe reflected light in only one direction
  • Analysing measurements without accounting for anisotropy can lead

to serious interpretation errors…

  • …but characterizing the anisotropy with multiangular measurements

can provide information on the structure of the target… something savanna ecologists are very interested in

slide-4
SLIDE 4

“ P h y s i c a l ” M I S R i n s t r u m e n t 9 a n g l e s x 4 b a n d s 3 6 n

  • n
  • r

e g i s t e r e d i m a g e s

I m a g e g r i d

“ V i r t u a l ” M I S R i n s t r u m e n t 9 a n g l e s x 4 b a n d s 3 6 c

  • r

e g i s t e r e d i m a g e s

S O M g r i d

W G S 8 4 e l l i p s

  • i

d E a r t h ʼ s s u r f a c e

SPACE OBLIQUE MERCATOR PROJECTION

MISR geo-location and angle-to-angle co-registration on Space Oblique Mercator (SOM) projection

Space Oblique Mercator projection minimizes re- sampling distortions 233 unique paths in 16-day repeat-cycle

  • f Terra orbit

MISR data is a pain in the ar**

9 cameras, 4 bands, weird HDF format, awkward projection

slide-5
SLIDE 5

A clever trick…

  • To save bandwidth, MISR only sends full

resolution (275 m) radiances for the 4 bands at nadir, and the Red band at all other angles

– The other bands are sent at 1.1 km resolution, except on special request

  • JRC algorithm reconstructs reflectances in all

bands at all angles and generates downstream products (e.g., FAPAR, LAI) at full resolution

  • This is unique to MISR – other FAPAR sources

are at 500 to 2500 m resolution

slide-6
SLIDE 6

Path 169 Block 110

Approximate Local Mode coverage

SANSA has the complete MISR archive for Africa east of Angola

slide-7
SLIDE 7

JRC RPV k parameter

  • bowl/bell shape of the anisotropy

k parameter, red band, Path 169, Orbit 17591, Block 110 k parameter, NIR band, Path 169, Orbit 17591, Block 110 Basalt- Granite contrast Bowl Bell

slide-8
SLIDE 8

Inversion

working backwards through the model

  • If we knew the surface

characteristics and the solar and observation angle, we could predict the photons reaching the satellite, using ray-tracing models

  • But we only know what the

satellite sees, and the solar and view angles

  • If we have enough view angles

we can work backwards to infer what the surface must be like

Pinty, B., T. Lavergne, R. E. Dickinson, J.-L. Widlowski, N. Gobron and M. M. Verstraete (2006) 'Simplifying the interaction of land surfaces with radiation for relating remote sensing products to climate models', Journal of Geophysical Research, 111, doi:1029/2005JD005952.

  • E. Bourderes
slide-9
SLIDE 9

Two-stream inversion package JRC TIP Surface albedo (soil + vegetation)

Surface albedo, VIS band, Path 169, Orbit 17591, Block 110 Surface albedo, NIR band, Path 169, Orbit 17591, Block 110

slide-10
SLIDE 10

JRC TIP Fraction of Absorbed Photosynthetically Active Radiation (FAPAR) and LAI

FAPAR, Path 169, Orbit 17591, Block 110 LAI, Path 169, Orbit 17591, Block 110 Phalaborwa gypsum dumps FAPAR=0, LAI=0 Good rains fell here

slide-11
SLIDE 11

Our goal: Detailed, long-term, accurate vegetation time-series

slide-12
SLIDE 12

Summary

  • 275 m resolution products of 8 to 16-daily FAPAR,

LAI, albedo and other surface parameters will soon be available for the lowveld, and later for all of Africa, for the decade 2000 to 2010+,

  • The accuracy and spatial resolution of the products

is unrivalled

  • They have application especially in climate

modelling, phenology, primary production modelling and change detection