Concepts in Precision Agriculture and Integrated pest Management

1
Concepts in Precision Agricultureand Integrated
pest Management
Jan 24, 2003

• Information compiled by
• Jason Kahabka, Cornell
• Lance Davidson, Cornell
• Art Lembo, Cornell
• Andy Roberts, Virginia Tech

SST
2
Outline

• Precision Agriculture and Precision IPM
• Talk about maps, spatial data and GIS
• GPS and scouting
• Quick overview of remote sensing
• Site -specific weather modeling
• Case study Look at GIS applications in IPM

3
Precision AgricultureWhat is it?
Precision Agriculture involves the application of
data acquisition/control systems and information
systems to land management and recognizes that
soil, crop, and pest-related processes are
variable in space and time within fields. The
use of Precision Agriculture is tied to new
technologies such as global positioning systems,
geographical information systems, and remote
sensing, and new statistical methods but it is
not strictly about TECHNOLOGY!
4
PA technology and knowledge
PA Technology refers to the hardware and software
that allows for the collection of information,
and control of crop management tools. PA
Knowledge refers to the integration of
information into a set of management tools that
allow for the optimum use of PA technology.
5
Information Agriculture (AKA Precision
Agriculture)
Information technologies are rapidly being
developed and employed for the purpose of
agricultural land management. Technologies such
as global positioning systems, geographical
information systems, and remote sensing, combined
with field-based data acquisition/control systems
allow for more informed decision making, more
efficient use of inputs, and better
record-keeping, thereby leading to more efficient
crop production and greater environmental
protection.
6
First, whats a map?
N
7
Data Models Raster vs. Vector
Vector GIS is composed of points, lines, and
polygons (areas). Raster GIS is composed of
rasters, or cells.

• Points
• Lines
• Areas

8
Discrete vs. Continuous Surfaces
Continuous surfaces are somewhat predictable.
There are an infinite number of locations that
have unique values - (e.g., ELEVATION) The
raster model is most suited to continuous data.
Discrete surfaces are not predictable. There are
a finite number of locations that have data -
(e.g., TAX RATES BY COUNTY) The vector model is
most suited to discrete data.
9
Add title
10
UNESCO Soils

• The Digital Soil Map of the World (version 3.0)
  was released in May 1994
• Scale of 15,000,000

11
Whats a GIS?

• A computer system capable of storing and using
  data describing places on the earths surface.

-or-
A set of interrelated functions that achieve
the Entry, Storage, Processing, Retrieval,
and Generation of spatial data
12
What questions can a GIS answer?

• Location What is at ...?
• Trends What has happened since...?
• Patterns What spatial patterns exist?
• Modeling What if...?

13
GIS Example Mapquest.com

• Problem Bad Grade in IPM

• Solution TP TAs house

• From phone book
• 421 S. Geneva St
• Ithaca, NY

14
GIS Maps Come From

• Digitizing
• Scanning
• Global Positioning Systems
• Remotely sensed imagery

15
GIS Functions

• Data Retrieval
• Map Generalization
• Map Abstraction
• Map Sheet Manipulation
• Buffer Generation
• Polygon Overlay
• Geocoding
• Dynamic Segmentation
• Network Analysis

16
Buffer Analysis

• A type of proximity analysis where a buffer zone
  is created to perform a search

• Point
• Circle
• Square
• Line
• Line Buffer
• Polygon
• Interior
• Exterior

17
Overlay

• Overlay is an operation in which sets of
  geographic regions are merged to form a new set
  of regions that the two initial sets share

18
How does it work?
Forest type
A
B
C

• One-theme layers are combined (overlayed) to
  produce multiple-theme layers.
• New information is derived from analysis of
  relationships among original data.

Ownership
1
2
3
4
C1
A2
A1
B2
B4
B1
C3
C4
Forest/ownership combination
19
GIS Functions Buffer, Clip, Reselect

purple
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
GPS-Global Positioning System

• The key to site specific management in the
  information age

25
GPS - How the System Works

• The Global Positioning System (GPS) is a
  worldwide radio-navigation system formed from a
  constellation of 24 satellites and their ground
  stations.
• The basis of GPS is triangulation from
  satellites.
• To "triangulate," a GPS receiver measures
  distance using the travel time of radio signals.
  To measure travel time, GPS needs very accurate
  timing which it achieves with some tricks.
• The system must correct for any delays the signal
  experiences as it travels through the atmosphere.

26
Differential Correction and SA

• GPS signals are no longer degraded but there
  are sill sources of error
• Differential correction is recommended for many
  PA applications
• Correction signals can come from satellite
  subscription, USCG navigation towers, or from the
  new WAAS system, where available.

27
(No Transcript)
28
Interpolation

• Values at discrete locations are point data.

• Continuous surfaces can be derived from point
  data through interpolation.

29
Inverse Distance Weight

• The influence of an input point on an
  interpolated value is isotropic. Since the
  influence of an input point on an interpolated
  value is distance related, IDW is not "ridge
  preserving"
• The best results from IDW are obtained when
  sampling is sufficiently dense with regard to the
  local variation you are attempting to simulate.
  If the sampling of input points is sparse or very
  uneven, the results may not sufficiently
  represent the desired surface

30
Kriging

• An advanced interpolation procedure that
  generates an estimated surface from a scattered
  set of points with z values.
• Assumes spatial correlation among data points
• Kriging is based on the regionalized variable
  theory that assumes that the spatial variation in
  the phenomenon represented by the z values is
  statistically homogeneous throughout the surface
  that is, the same pattern of variation can be
  observed at all locations on the surface.

31
Using a Semivariogram to determine an appropriate
sampling grid
150
Variance
50 100 150 200 250 Distance (m)
32
Theissen
33
Inverse Distance Weighting
34
Kriging
35
An example soil matric potential
as a covariate in analysis of soil-borne diseases
36
Remote Sensing

• Remote Sensing The techniques for collecting
  information about an object and its surroundings
  from a distance without contact
• Components of Remote Sensing
• the source, the sensor, interaction with the
  Earths surface, interaction with the atmosphere

37
The First Application of Remote Sensing
38
Remote Sensing Information

• High-resolution aircraft and satellite-based
  remote sensing information is now available at
  reasonable cost that is
• multi-spectral (hyper-spectral)
• digital
• georeferenced
• fixed-wing aircraft
• satellite

39
Aerial Image Analysis
Bare soil bands ? PCA
Vegetation bands ? NDVI
40
Short-Wave Reflectance Patterns
Healthy crop
r
Stressed crop
Dry / low OM soil
Wet / high OM soil
1000
100
UV blue green red NIR
---------- visible --------
WAVELENGTH (nm)
41
Sensors take multiple bw images at discrete
wavelengths
Remote Sensing
Spectral signatures
42
Principle of Reflectance
43
False Color Composites
B W images from different bands are assigned
colors combined.
44
Vegetative Indices
45
MRLChttp//edc.usgs.gov/glis/hyper/guide/mrlc

• The Multi-Resolution Land Characteristics (MRLC)
  project was established to provide
  multi-resolution landcover data of the
  conterminous United States from local to regional
  scales.
• Objective to develop a national 30-meter
  landcover characteristics data base using Landsat
  thematic mapper (TM) data. This is a cooperative
  effort among six programs within four U.S.
  Government agencies

46
Satellite Imagery for Plant Disease Detection
Clubroot on a cauliflower crop
47
Spatial Resolution for Satellites

• Landsat MSS 80 meter
• Landsat TM 30 meter
• SPOT 10 meter
• IRS 6 to 28 meter

48

• According to today's yield monitor manufacturers,
  most users should obtain accuracy within /- 3
  percent, if the system is properly installed,
  maintained and calibrated. Items that operators
  must be conscious of and attend to for good
  results can be summarized as follows (PPI)
• Proper calibration of the mass-flow sensor using
  multiple loads acquired according to the
  manufacturer's recommendations.
• Inspection of the system sensors, particularly
  those affected by crop conditions, during the
  harvest.
• Verification and, if necessary, calibration of
  the ground speed sensor.
• Verification and calibration of moisture and
  temperature sensors.
• Correct entry of the operating information such
  as crop type, field, and header width for each
  field into the system console.
• Proper use of the software to extract and process
  the yield data.

49
John Deere CropTracer TM
Consumers, processors, grain buyers, and trade
partners are demanding more accountability from
producers. STEP 1 Capture Essential Data
From primary tillage practices to harvest
information, John Deere GreenStar precision
farming systems let you electronically monitor
and record nearly all aspects of your crop
practices and field conditions. STEP 2
Validate Your Practices CropVerifeye offers a
long list of field audits and services that trace
the genetic integrity and identity of your
agricultural products. They can also provide
identity-preserved training programs. STEP 3
Manage the Information VantagePoint Network is
the perfect place to store, organize, and analyze
farm data. It provides a secure method for you to
provide this valuable information to trusted
suppliers, contractors, and advisors. Why is
This Level of Documentation Necessary?
CONSUMERS WANT IT. Recent product containment
issues have consumers worldwide focusing on food
safety. They are beginning to demand labeling of
food products - particularly from genetically
modified crops. TRADE PARTNERS DEMAND IT. As
countries continue their export trading, there's
a huge need for an information system that can
back up the quality and integrity of grain and
food products with third-party verification.
50
Penetrometer 24.5-35.0 cm Chisel Till vs.
Zone Till
Late
Early
Early
Late
Root Limiting
Less Strength
More Strength
51
Global Positioning

• Determines the geographical location of the
  equipment (georeferencing)
• In IA, global positioning systems are used
  jointly with field data acquisition (e.g., yield
  measurements) or control (e.g., fertilizer/lime
  application) systems, or sampling equipment
  (e.g., soil samples, penetrometer measurements).

52
Using GIS to apply insecticides is possible but
perhaps less cost-effective than its use with
either herbicides or fungicides. Insects are much
more vagile and within field distribution may
vary year to year. However, certain trends in
insect population distribution can be easily
demonstrated using GIS.
http//www.colostate.edu/Depts/IPM/tmp/gis_ag/ipmg
is.html
53
Weather modeling
Computers also have made it possible for
agricultural service companies to deliver
site-specific weather information, which can help
growers improve insect and disease management as
well as other production practices. SkyBit, Inc.,
of Boalsburg, which recently introduced such a
service to producers, takes raw National Weather
Service data, then uses computers and information
technologies to develop customized weather
forecasts, which can be tailored to each farmer's
needs. "Our company provides these forecasts to
growers across the country," says Joseph Russo,
president of SkyBit. "We can produce two-day
reports that forecast the weather for a single
farm on an hourly basis." By entering this
information into pest development models based on
weather observations, SkyBit also can forecast
when insect and disease problems will develop on
individual farms. "Many IPM programs use weather
data to determine when crops are likely to be
threatened by various insects and diseases, and
when insects are most likely to be controlled
successfully," Russo says. "For instance, tufted
apple bud moth is susceptible to most
insecticides for only a brief period after the
larvae hatch. By evaluating weather conditions at
the farm level, SkyBit can help growers know
exactly when the larvae are most likely to appear
in their orchards."
54
(No Transcript)
55
Everything Is Related to Everything Else, but
Near Things Are More Related Than Distant
Things - Waldo Toblers First Law of Geography
56
Using a Semivariogram to determine an appropriate
sampling grid
150
Variance
50 100 150 200 250 Distance (m)
57
Depth to Basal Till
deeper
shallower
58
1999 SARE 03pH
Here the pH along the western edge of the field
is 5.1 -5.7
59
Sources of information

• Soil survey
• Soil samples
• Aerial images
• Yield monitor data
• My work attempts to evaluate how useful each of
  these sources of information are in defining new
  management zones. Analysis is based on 3 years of
  data for 5 fields located in Onondaga and Seneca
  counties (central NY).

60
Soil survey
Detailed soil series locations obtained from
SSURGO or by digitizing Soil Survey maps.
61
Soil samples
Soil sample data was interpolated using
geostatistics (kriging or inverse distance
weighting). This was done for pH, P, K, OM and
NO3.
62
Aerial images
Aerial images taken by Emerge. 2 images per
season one of bare soil (June), the other of
vegetative cover (August). Images have 3 bands
green, red and near-infrared.
63
Coordinate systems

• Geographic space - The earth is spheroid so
  locations are measured in degrees.
• Latitude / longitude
• Projected space - round surfaces must be
  projected to make them display on a flat map
• UTM (universal transverse mercator) measured in
  meters

64
Map Datum

• Measurement convention or mathematical model that
  describes the shape of the world.
• GPS based on World Geodetic System 1984 (WGS
  84)
• USGS quad maps based on North American Datum 1927
  (NAD27)
• Future maps will be based on (NAD83)
• Other countries have their own datums

65
Projections

• GPS data will need to be projected before
  overlaying a map
• Projections can be a source of frustration for
  new users of GIS.

66
Projections

• A function or transformation which relates
  coordinates of points on a curved surface to
  coordinates of points on a plane.

67
UTM

• Angles are preserved, but distances away from the
  equator become progressively distorted. (for
  example South America is nine times larger than
  Greenland.

• Over small areas the shapes of objects will be
  preserved.
• A line drawn with constant orientation (e.g. with
  respect to north) will be straight on a conformal
  projection (a rhumb line).
• Parallels and meridians cross each other at right
  angles (note not all projections with this
  appearance are conformal).
• Generally, areas near margins have a larger scale
  than areas near the center (i.e. Greenland in
  Mercator projection).

Source http//math.rice.edu/lanius/pres/map/mapp
ro.html
68
The 60 UTM zones
69
UTM zones (cont)
Zones converge like lines of meridian but square
grids are easier to use.
70
UTM zones (cont)

• Coordinates are measured in Northing and Easting
• UTM uses a regular grid with a False Easting
• Center is 500,000 E

71
GPS Software

• Most have common features and a similar interface
• Manufacturers
• Garmin
• Trimble
• Magellan
• Many yield monitor systems have other GPS
  utilities that can be used for mapping

72
Yield monitor data
Yield data from a combine harvester equipped
with a DGPS unit and yield monitor.
73
Integrated analysis

• Analysis techniques
• Correlations
• Stepwise regression
• PCA or Factor Analysis
• Spatial regression

74
Kriging

• If spatial structure is apparent then Kriging can
  be used to interpolate a map, assigning values to
  areas in between sample points.

75
Musgrave Farm - Aurora, NY
76
All Harvest Plots - Field Z
77
Soil Test P
78
An Example of GIS in IPM
Slow the Spread (STS), initiated to manage the
spread of Gypsy moth populations Example courtesy
of Andy Roberts, Virginia Tech
http//www.fs.fed.us/ne/morgantown/4557/gmoth/
79
Slow The Spread...Slow What Spread???
80
Biological basis of STS

• Implemented immediately ahead of the advancing
  front.
• Detect and suppress isolated gypsy moth colonies
  before they grow and coalesce.

WV
VA
NC
WV
VA
KM
0
50
81
Montgomery Co.,Virginia
82
The Problem

• Where to apply?
• Landcover is deciduous forest.
• Interpolated catch is 1.
• What to apply?
• Bt within 100 meters of streams.
• Dimilin elsewhere.
• When to apply?
• Below 2500 feet elevation first.
• At or above 2500 feet second.

83
Roads
Streams
Elevation
Land cover
84
Step One Where is the entire treatment area?

• Catch 1.
• Land cover is deciduous forest.

85
1994 Survey sites
Positive sites are in red.
Interpolated catch surface
86
Landcover / catch overlay
87
Gypsy moth treatment area

• catch 1
• LC forest

88
Step Two Where will the two treatments types be
applied?

• Bt within 100 meters of streams
• Dimilin elsewhere

89
Detail of streams within the treatment area
90
Streams within treatment area

• 100 meter buffers

91
Location of the the two treatment types
92
Step Three When will the treatments be applied?

• Elevations below 2500 feet first
• Elevations above 2500 feet second

93

• Up close
• Elevation in 500 ft contours
• Location of the treatment area
• (Notice raster vector data)

94
The solution!

• Deciduous forest
• Catch 1
• Bt 100 meters from streams
• Low elevations first
• High elevations second

95
The solution!

• Treatment acreages
• Catch Land Cover Treatment Elevation Blocks
  Hectares
• 1 Forest Dimilin Low
  3 220
• 1 Forest Dimilin High
  2 19
• 1 Forest Bt Low 3
  52
• 1 Forest Bt High 1
  1

96
The conclusion!
Some CoursesCSS 411. Resource Inventory
MethodsCSS 420. Geographic Information
SystemsCSS 620. Spatial Modeling and
AnalysisCSS 465. Global Positioning
Systems People in the know at Cornell Art
Lembo Ann-Margaret Esnard
Old geographers never die, they just lose their
bearings...