Landscape Position Zones and Reference Strips

1
Landscape Position Zones and Reference Strips

• Larry Hendrickson

2

• Landscape Position Zones

3
Landscape Position Zones (LSP)

• Extracting landscape position (LSP) from
  elevation data
• RTK elevation data is becoming widely available
• Elevation derivatives work well for post-mortem
  analysis
• Elevation itself isnt usually a good means for
  classifying site conditions
• Developed a method to extract LSP from elevation
  data by comparing elevation of each pixel with
  its neighbors

4
W Nebraska Center Pivot field
W Nebraska Center Pivot field
5
LSP Zone Polygons over Corn Yield
Shoulder
Toeslope
6
Yield variability related to LSPAcross W NE
fields
Toeslope
Shoulder
7
Yield variability related to LSPAcross years
Toeslope
Shoulder
8
Iowa field just after planting
9
Iowa field just after planting
5 4 3 2 1
10
Iowa Corn Yield
May-June Rainfall
6.1
7.8
7.6
9.4
13.2
14.9
Toeslope
Shoulder
Normal May-June rainfall is 9.5 inches Data from
Jaynes and Kaspar
11
Iowa Soybean Yield
May-June Rainfall
7.3
1.6
16.8
12.8
14.8
Toeslope
Shoulder
Normal May-June rainfall is 9.5 inches Data from
Jaynes and Kaspar
12
LSP Comparisons to Soil Survey Maps
13
Landscape Position Zones

• LSP is an intuitive characteristic
• Water runs downhill
• Drier on shoulders, wetter on toeslopes
• Often better relationship to yield than soil
  conductivity
• In regions where topography was the primary soil
  forming factor

14

• Reference Strips

15
Role of Reference Strip in N Rate Decisions

• Studies conducted by Sawyer in 2005 and 2006
• Focus on reference strip observations from these
  studies
• Patterns observed with aerial imagery
• Compare N stress outcomes with and without
  reference strips

16
Approach

• 3 reference strips in each field with 240 or 270
  N/acre
• Aerial images were taken between V10 and V14
• Calculated GNDVI from aerial images
• SPAD readings were taken within 3 days of images

17
Field 1
240 N
GNDVI
GNDVI with upper 10 in blue
18
Field 2
19
Field 3
20
Imagery patterns

• Considerable variability in N stress within and
  between reference strips
• High GNDVI values were found within all
  fertilized (60-180N) strips, but not within zero
  N strips
• High GNDVI values were also found in other parts
  of all fields examined
• High GNDVI values were often impacted more by
  soil variability than by reference strips

21
(No Transcript)
22
Data Analysis

• Extracted underlying GNDVI from each SPAD
  location
• Related relative SPAD, relative GNDVI, and GNDVI
  relative to best area to yields underlying
  these points
• Relationships across 8 studies in 2006

23
R2 0.44
24
R2 0.53
25
R2 0.52
26
Summary

• Imagery (and presumably sensors) were more
  closely related to yields than SPAD values across
  fields
• GNDVI relative to best areas was equivalent to
  using a reference strip
• Reference strips were unnecessary for N decisions
  made after V10 in fields that had received
  significant rates of fertilizer earlier in the
  season (in Iowa)
• Likely cant extrapolate these results to other
  regions or earlier N applications

27

• Sensors and Soil Zones

28
Landscape position on N requirement
Shoulder Toeslope
Yield Potential
N Mineralization
N Loss Potential
Patterns of N stress observed will depend upon
which factor dominates in a particular region or
year
29
Estimating N Requirement
Sensor Observes Differential Residual
N Mineralization Losses of soil and fertilizer
N Crop stand and growth patterns
All are impacted by LSP or SC zones
Early N
Soil N
PL
Sensing
30
Estimating N Requirement
Sensor Observes Differential Residual
N Mineralization Losses of soil and fertilizer
N Crop stand and growth patterns
Estimate Differential Losses of soil and
fertilizer N Mineralization Crop stand and growth
patterns
Early N
?
Soil N
PL
Sensing
31
Estimating N Requirement
Sensor Observes Differential Residual
N Mineralization Losses of soil and fertilizer
N Crop stand and growth patterns
Estimate Differential Losses of soil and
fertilizer N Mineralization Crop stand and growth
patterns
Early N
Both are often impacted by LSP or SC zones
?
Soil N
PL
Sensing
32
Linking Soil Zones to Sensors

• Soil zones are likely quite stable over time
• Opportunity to re-evaluate existing sensor data
  obtained in large field studies by acquiring soil
  zones
• Incorporate soil zones in future sensor
  evaluations
• Soil zones may be useful for
• Early season applications (in fields that behave
  consistently across years)
• Modifying algorithms to reflect differential N
  requirements
• Directing initial pass with sensors

33
Use of zones to help select best areas

• Imagery has advantage over sensors in that it
  provides data for entire field
• Selection of best area during initial pass may
  offer sufficient assessment
• Seems logical that best area in field will be
  in extreme soil zones (wettest/driest,
  darkest/lightest soils)
• First pass in field should be through areas with
  most extreme soil conditions

34
New 2510H Applicator
Extends Sidedress Window Clearance allows
application to 30 corn High Speed10 mph NH3 or
UAN
35
(No Transcript)