Project 4'3:

1
Project 4.3

• A Near Real Time whole Farm Package integrated
  remote sensing technologies for improved farm
  management

2
Project objectives

• To deliver an operational, NRT, easy to access,
  cost effective whole farm package of pasture
  and crop
• Growth rate
• Biomass
• Pasture quality
• that can be used by producers to make better
  tactical strategic decisions at paddock farm
  level.

3
Specifics

• Pasture
• Develop a capacity to estimate paddock-level
  pasture quality characteristics (botanical
  composition, nitrogen, fibre, carbohydrate, in
  vitro dry matter digestibility, organic matter
  and lignin/cellulose) at monthly intervals in the
  winter rainfall dominant Mediterranean region of
  WA
• Crops
• Incorporate RS (plus soil weather stations
  networks) data to DAWAs STIN crop growth model,
  so that a better spatialization of crop growth
  status and yield forecasting can be done at farm
  level.

4
What other crop/pasture RS-based systems are
available?
5
Other RS crop monitoring systems

• Crop View
• AgrowatchTM
• Geotechnologies (USA)
• Crop View (USA)
• SkyPlan

6
Crop View

• CropView identifies spatial patterns in crop
  paddocks. (Spot based)
• Early crop view (3 images)
• Differences in crop growth 608 wks post sowing,
  weed detection?
• Differences in crop growth late tiling prior to
  booting gt fertilizers
• Harvest planning (2 images)
• Provide Yield estimates gt (doesnt account for
  rainfall, or effects of soil properties that show
  later on the growing season)
• may enable measurements of crop damage
• Display pest and disease damage to canola
  cereals.
• Each pixel represents the average crop vigour
  reflectance of an area of 20x20 m

7
Crop View..

• Crop vigour is measured with Vegetation Index
  (VI)that evaluates crop health and biomass as
  indicated by chlorophyll activity.
• Images delivered via email, 2-3 days of satellite
  overpass
• Retrospective gt NRT, not predictive
  capabilities.
• Image interpretation largely left to farmers gt
  farmers must provide their ground truth.
• Difference images gt crop change over time.

8
AgrowatchTM

• Internet delivery (within 3 days of satellite
  pass)
• Based on Quickbird, Spot, Landsat
• Products
• HR (Quickbird) MR (Spot 2-5) HRS (Quickbird Pan
  MS) BDRF corrected
• Deliver color coded field maps (qualitative)
• Crop status (NDVIg vegetation map) gt amount of
  green vegetation
• Soil condition (Soil map -gt soil brightness
  index)
• Rate of crop change (difference image remove
  variation in veg responses due to influence of
  soils (T2-T1) determine rate of change in crops
• ScoutAide how fast a plant is growing or dying
• Images are BDRF corrected only.

9
AGROWATCH processing
Similar to existing Perth services SpecTerra
Services.
10
Skyplan

• Retrospective, provides areas of variation
  within paddock using historical data.
• Data Landsat TM
• Enables farmers to isolate good from bad yielding
  areas.
• Yield maps based on biomass from imagery.

11
Upper Midwest Aerospace Consortium (UMAC)

• Delivery Internet
• Uses
• MODIS gt NDVI gt change images.
• Landsat ETM, IKONOS, Hyperion, Aster, airborne
  multispectral for individualised needs of
  members.
• Provide training to farmers on GIS, image
  interpretation.
• AOI for data distribution

12
UMAC applications development

• Monitoring wheat with weekly NDVI (vs historical
  data)
• Stress detection monitoring field conditions
  (airborne-based)
• Zoning VR applicatons of nitrogen
• Selecting sugarbeet areas for payment in kind
• Effectiveness of fungicide application
• Detection of insect infestation
• Spray drift damage assessment
• Drainage deficiency/inundation.

Something is wrong what, why?
13
Geotechnologies (Denver, USA)

• Web-based online GIS
• Monitoring crops, for growth, irrigation
• Data uses Ikonos
• Products coloured change images showing changes
  in crop growth
• System allows interactive tracking, can do
  automatic border detection as tractor moves
  across field

14
Crop View EarthSat (USA)

• The intent of this information is to show which
  areas of your fields are under stress
• Uses Landsat-7

15
What is good from existing?

• New indices are available
• Provide qualitative indication of within-
  between paddocks variation
• Comparison against historical avg. gt paddock may
  experience problems
• Spatial assessment of damage/ effect of
  management measures.
• Monitoring within growing season

16
What existing systems lack?

• Crops none account for water effect (e.g.
  rainfall) or soil variations in yield
  estimations
• Many are retrospective not real NRT predictive
• Many offer qualitative indications of crop
  vigour, growth rate, etc, based on spectral
  indices and multidate image comparison.
• A real indication of costing (e.g. Agrowatch)
• In Australia WATER is one of the main limitating
  factors

17
What we propose?Crops
STIN
MODIS corrected (project 4.1)
Deliver paddock information about crop growth
rate (Indices-biomass) (MODIS-based NRT maps of
Crop Biomass)
Delivers Soil moisture availability Yield
estimation
For AOI
Overtime develop relation Yield forecast
biomass (RS-modelled) gt Calibration of STIN.
Y
For AOI
R
Ideal Yield (STIN)
Biomass/Yield
NRT CGR, plus comparison against ideal yield
(modelled including rainfall, soils)
Biomass as derived from MODIS (anomalies)
Growing season
Farmers to make tactical decisions
18
(No Transcript)
19
(No Transcript)
20
Advantages

• Scientific understanding for the use of MODIS
  corrected products for paddock-scale monitoring
  of crops/pastures
• Higher resolution products (like AgWatch, UMAC)
  selective, as demander by user community (e.g.
  based on Spot, Quickbird, Landsat, or airborne
  multispectral imagery).
• These NRT products should allow manager to act on
  information that tells what is happening NOW, not
  last season/year(s).

21
Pastures

• Maps of Pasture quality at paddock level

22
The package
Crops
Pastures

• MODIS-based NRT PGR
• MODIS-based maps of 7 days forecast PGR
• HSR SPOT-HRG Landsat TM maps FOO
• MODIS based maps of FOO

• MODIS-based NRT maps CGR
• MODIS based forecasting maps yield
  (STIN-satellite)
• HSR-based NRT predictive maps of yield
  (STIN-satellite)

• Pasture quality maps (airborne satellite based)

To be developed within 4.3
To be provided by Pastures from Space
23
(No Transcript)