Crop Biomass Estimation with NRT MODIS Data and Early Results of a Total Crop Area Algorithm

1

• Crop Biomass Estimation with NRT MODIS Data and
  Early Results of a Total Crop Area Algorithm
• Integrated remote sensing technologies for
  improved farm management
• Dr Andrew P. Rodger
• arodger_at_agric.wa.gov.au
• Curtin University of Technology
• Department of Agriculture Western Australia

2
What are we Attempting to Measure and Where?

• What?
• Total crop area on a state, shire, and farm level
• Soil end-members from MODIS data collected over
  the Western Australian wheat-belt.
• Farm level fractional crop coverage
• Potential crop yield at the farm level
• Where?
• Western Australian wheat belt in general (May
  2005-Present)
• Merredin Research Station (DAWA) (May
  2005-Present)
• Glenvar Farm at Wongan Hills, WA. (May 2005-
  Present)
• Additional farms at Wongan Hills as the season
  progresses

3
Merredin and Glenvar
4
Automatic Soil End-Member (AUTOSEM)
softwareSelection Using MODIS Imagery

• Uses the first 5 MODIS bands (459-1250nm)
• Uses a time series of MODIS reflectance imagery
• Each MODIS reflectance spectrum is compared to a
  known soils spectral library (ASTER) and the sum
  of the differences from each of the 5 bands
  recorded
• For a given pixel the MODIS spectra with the
  smallest sum difference is assumed to be probable
  soil
• The error of fit of the MODIS spectra to the
  ASTER library (at present) spectra is then
  compared to a predefined tolerance level
• All spectra having an error of fit greater than
  the tolerance level are classed as non-soil

5
AUTOSEM _at_ Merredin Initial Results
6
MODIS Wavelength versus Zeiss Spectrometry
AUTOSEM Reflectance

• Good agreement between MODIS-derived soil spectra
  and those measured at Merredin
• Comparable variation in the first 4 MODIS bands.

7
Preliminary Results (1)
Fractional Analysis Model Using a Zeiss
Vegetation End-Member and the AUTOSEM Sharpees
Farm Using the AUTOSEM in combination with a
Zeiss-derived vegetation end-member yields the
following time series of fractional vegetation
cover. These results do not distinguish between
crop species, remnant vegetation or natural
vegetation.
8
Preliminary Results (2)
While the overall trend is what we expect, the
confidence in the vegetation end-member is not
high since we have yet to gather a representative
sample of crop end-members.
9
Preliminary Results (3)
Potential STIN yield prediction for Wongan Hills
as of the 22/7/05 is 2.305 (T/ha). The potential
mass of the crop Mc can be estimated at any point
in time as McVF6.252.305, where VFvegetation
fraction and the factor 6.25 is the size of a
MODIS 250m pixel in hectares. The assumption is
that the peak tonnage corresponds to the end of
year yield. This is no different than the
assumption of mid-season NDVI being the best
indicator of yield.
10
Total Crop Area from Temporal MODIS NDVI
11
The Team
The Partners

• Project Leader
• Professor Graciela Metternicht,
• Dept. of Spatial Sciences, Curtin Univ.
  Technology
• Research Team
• Dr Andrew Rodger Dr Tom Schut, Research
  Associates, Curtin Univ. Dept Agriculture WA
• Mr Greg Beeston, Dept. Agriculture Western
  Australia
• Dr Stephen Gherardi, Dept. Agriculture Western
  Australia
• Dr Miles Dracup, Dept. Agriculture Western
  Australia
• Mr Damien Shepherd, Dept. Agriculture Western
  Australia
• Dr Richard Smith, Dept of Land Information
  Western Australia
• Dr Dave Henry and Dr Asoka Edirisinghe, CSIRO