Download On the Go Soil Sensing - Lecture Slides | AGEN 431 and more Exams Engineering in PDF only on Docsity!
On- On-thethe--Go Soil SensingGo Soil Sensing
AGRO/MSYM/AGEN 431AGRO/MSYM/AGEN 431
Viacheslav I. Adamchuk Viacheslav I. Adamchuk
Biological Systems Engineering DepartmentBiological Systems Engineering Department
University of NebraskaUniversity of Nebraska--LincolnLincoln
October 7, 2008October 7, 2008
Problem Statement Problem Statement
- The assessment of soil variability is one of the most important steps in site-specific management
- Conventional means to attain soil variability data are incapable of accurately identifying spatial inconsistency within a production field at an economically feasible cost
- There is a need to develop equipment for mapping soil attributes on-the-go
Sensor Use ApproachesSensor Use Approaches
Map-Based Approach
Integrated Approach (Real-Time with Supplemental Base Map)
Real-Time Application
Nature of Proximity SensingNature of Proximity Sensing
Listen
Taste
Smell
Touch
Look
OnOn--thethe--go Soil Sensorsgo Soil Sensors
Electrical and
Electromagnetic
Acoustic
Mechanical Electrochemical
H + H +^ H +
H + H +
Pneumatic
Optical and
Radiometric
Electrical end Electromagnetic Sensors Electrical end Electromagnetic Sensors
Electrical Conductivity/Resistivity Sensors
ECa
EC (^1) EC (^2) EC (^4) EC (^3)
EC (^5)
Soil Soil
Galvanic Contact Resistivity Method
Electromagnetic Induction Method
Capacitively-Coupled Resistivity Method
- Soil type/texture
- Salinity
- Water content
- Organic matter content
- Depth variability
- Soil pH / nitrate content
- Volumetric water content
- Soil type/structure
- Salinity
Dielectric Sensors
Magnetic Sensors
- Subsurface soil impurities
- Iron
Galvanic Contact Resistivity MethodGalvanic Contact Resistivity Method
I A U
NB M
Current flow
Equipotentials
Veris Technologies, Inc. (Salina, Kansas) http://www.veristech.com
Veris®^ 3100 and MSP (0.3 and 0.9 m)
Geocarta (Paris, France) http://www.geocarta.net
Geocarta ARP (0.5, 1, and 2 m)
Crop Tehchnologies, Inc. (Spring, Texas) http://www.soildoctor.com
Soil Doctor ®^ System (real-time approach)
Electromagnetic Induction Method Electromagnetic Induction Method
Receiver
Primary field Soil Eddy currents
Transmitter
Secondary field
Geonics Limited (Mississauga, Ontario) http://www.geonics.com
Geonics EM- horizontal – 0.75 m vertical – 1.5 m
Dualem, Inc. (Milton, Ontario) http://www.dualem.com
DUALEM – 1S co-planar – 0.4 m perpendicular – 0.95 m
CapacitivelyCapacitively--Coupled ResistivityCoupled Resistivity
MethodMethod
Capacitor analogue Metal shield as acapacitor plate
Soil as a capacitor plate
Insulation as dielectric material
Coaxial cable
Soil
Transmitter
Inner wire
Geometrics, Inc. (San Jose, California) http://www.geometrics.com
Geometrix OhmMapper TR
Example 1Example 1
Electrical Conductivity MapElectrical Conductivity Map
Improved Soil Type Separation
Soil Survey EC Map
Example 2Example 2
Electrical Conductivity Map Electrical Conductivity Map
EC Map
Low Yielding Area
High Yielding Area
Yield Map
Example 3Example 3
Electrical Conductivity Map Electrical Conductivity Map
Shallow EC Deep EC
Excessive manure application^ Area of excessive manure application
Shallow/Deep EC
VIS/NIR SpectrophotometerVIS/NIR Spectrophotometer
Predicted Carbon Measured Carbon Sapphire Window Veris Technologies, Inc. (Salina, Kansas) http://www.veristech.com
Traveling Spectrophotometer Traveling Spectrophotometer
CCD camera
Optical fiber for visible reflection
Soil flattener Soil surfaceillumination
Optical fibers for illumination
Penetrator tip
Shank NIR thermometer
Laser displacesensor Optical fiber for NIR reflection
Travel direction
Ground surface
Tokyo University of Agriculture and Technology (Tokyo, Japan) Load Cell EC Electrode
Mechanical Sensors Mechanical Sensors
Cantilever beam sensors
Direct load sensors
Single-tip horizontal sensors
Multiple-tip horizontal sensors
Vertically oscillating sensors
Vertically- operated cone penetrometers
Soil profile sensors
Tine-based sensors
Tip-based sensors
Soil strength sensors
- Soil mechanical resistance
- Soil compaction
- Water content
- Soil types
- Depth of hard (plow) pan
Vertically actuated sensors
Bulk soil strength sensors
Draft and vertical force sensors
Strain Gauges
Soil Mechanical Resistance MappingSoil Mechanical Resistance Mapping
Tool Bar
Travel Direction
Purdue University (West Lafayette, Indiana)
Discrete Depth Profiling ToolsDiscrete Depth Profiling Tools
UC-Davis (Davis, California)
Three CuttingThree Cutting BladesBlades
UNL (Lincoln, Nebraska)
University of Missouri (Columbia, Missouri)
Load CellLoad Cell ArrayArray
ExampleExample
Soil Mechanical Resistance Map Soil Mechanical Resistance Map
Soil Mechanical Resistance Map (20-30 cm)
Yield Map
Compacted area Old roads
Integrated Soil Physical Properties Integrated Soil Physical Properties
Mapping SystemMapping System
Two wavelengths soil reflectance sensor
Soil mechanical resistance profiler with an array of strain gage bridges
Capacitor-based sensor
UNL
(Lincoln, Nebraska)
Vertical Blade with Strain Gage ArrayVertical Blade with Strain Gage Array
UNL (Lincoln, Nebraska)
Discrete Model
Soil surface
Travel direction
Strain gages
Polynomial Model
Apparent soil surface
Apparent Soil Profiles Apparent Soil Profiles
Plot B (disked)
0
5
10
15
20
25
-0.8 -0.4 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.
Soil mechanical resistance, MPa
Relative depth, cmISPPMS - pass 1Cone - pass 1 ISPPMS - pass 2 Cone - pass 2
Plot D (chiselled and disked)
0
5
10
15
20
25
-0.8 -0.4 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.
Soil mechanical resistance, MPa
Relative depth, cm
ISPPMS - pass 1 Cone - pass 1 ISPPMS - pass 2 Cone - pass 2
Integrated Load ComparisonIntegrated Load Comparison
1
3
5
7
9
1 1
1 3
0 2000 4000 6000 8000 10000 Overall integrated load (ISPPMS), N
Tillage plot
Center row Wheel row
A
F
E
D
C
B
pass 1
2
1
2
1
2
1
2
1
2
1
2
1
3
5
7
9
1 1
1 3
0 2000 4000 6000 8000 10000 Overall integrated load (cone penetrometer), N
Tillage plot
Center row Wheel row
A
F
E
D
C
B
pass 1
2
1
2
1
2
1
2
1
2
1
2
Instrumented Blade Cone Penetrometer
B disked E double disked C no-till with cultivation F no-till w/o cultivation
A plowed and double disked D chiselled and disked
Tillage Plot Treatment Tillage Plot Treatment
Vertical Blade with Strain Gage ArrayVertical Blade with Strain Gage Array
UNL (Lincoln, Nebraska)
Discrete Model
Soil surface
Travel direction
Strain gages
Polynomial Model
Apparent soil surface
Linear Model
Soil Mechanical Resistance Soil Mechanical Resistance
Profile
Average
5 – 30 cm depth
Increase
with Depth
Sample collection for calibration Cleaning
ISFET Electrode
Water jet ~50 g soil core
Mixing
Add 20 ml DI H 2 O
Automated Soil Testing Automated Soil Testing
Shank
Soil cutters
Coring tube
Purdue University (West Lafayette, Indiana)
Soil/Buffer pH Mapping OnSoil/Buffer pH Mapping On--thethe--GoGo
The University of Sydney (Sydney, Australia) JTI (Uppsala, Sweden)
Waived sampling disc
Soil preparation and analysis unit
Automated Soil pH Mapping SystemsAutomated Soil pH Mapping Systems
US Patent No. 6,356,
Purdue University (West Lafayette, Indiana)
Soil Sampling Mechanism
Travel Direction
Water Supply
Water Nozzle
Sensors Output
Soil Shank
Removable Plates pH Sensor
Air Supply Air Cylinder
Soil Sample
Sampling Platform
5 mm
Soil pH Measurements OnSoil pH Measurements On--thethe--GoGo
Distance from the W est End, m
Soil pH
Data Set 1 Data Set 2
Laboratory
Mobil Sensor Platform (MSP)Mobil Sensor Platform (MSP)
Veris Technologies, Inc. (Salina, Kansas)
http://www.veristech.com
EC Surveyor 3150
Soil pH Manager
Direct Soil MeasurementDirect Soil Measurement
Purdue University (West Lafayette, Indiana) Veris Technologies, Inc. (Salina, Kansas) UNL (Lincoln, Nebraska)
Water Nozzle
Soil Sampler
Ion-selective Electrodes
ExampleExample
Soil pH MappingSoil pH Mapping
Soil pH Maps of
a Kansas Field
On-the-Go Mapping Conventional 1 ha Grid Sampling
Directed Soil Sampling
Evaluation Fields Evaluation Fields
27-84 acre fields 12-34 grid samples (0.3-0.5 samples/acre) 250-598 MSP measurements (4-11 measurements/acre) 5 calibration samples & 5 validation samples
WI1 Silt loam 18% 6.66 (0.47) 3.22 (1.08)
OK1 Loamy fine sand 2% 6.16 (0.64) 0.96 (0.99)
NE1 Silty clay loam 11% 5.95 (0.84) 25.86 (4.97)
KS2 Silty clay loam 3% 6.62 (0.68) 16.49 (4.6)
KS1 Silt loam / silty clay loam 6% 5.34 (0.27) 3.17 (1.00)
IL2 Loam / clay loam 2% 6.52 (0.86) 14.88 (3.66)
IL1 Loam / clay loam 2% 6.28 (0.41) 11.44 (2.22)
IA1 Loam / silty clay loam 5% 5.18 (0.77) 9.26 (5.58)
Field ID Textural range Max slope Lab pH* EC (mS m -1) *
Mapping AlternativesMapping Alternatives
Universal MSPpH = InterceptUniversal + SlopeUniversal ⋅ MSP pH
Adjusted MSPpH = InterceptField − specific + SlopeField − specific ⋅ MSP pH
Shifted MSPpH = ShiftField − specific + MSP pH
On-the-Go Sensing Soil Sampling
MSP pH
Universal MSP pH
Directed Sampling
Adjusted MSP pH
Shifted MSP pH
Field Average
Interpolated Grid Map
Soil pH Maps Evaluation Soil pH Maps Evaluation
4
5
6
7
8
9
4 5 6 7 8 9 Grid Sampling pH
Lab pH
IA1IL IL2KS KS2NE OK1WI 1:1 lineReg 4
5
6
7
8
9
4 5 6 7 8 9 Field Average pH
Lab pH
IA1IL IL2KS KS2NE OK1WI 1:1 lineReg
4
5
6
7
8
9
4 5 6 7 8 9 Shifted MSP pH
Lab pH
IA1IL IL2KS KS2NE OK1WI 1:1 lineReg 4
5
6
7
8
9
4 5 6 7 8 9 Adjusted MSP pH
Lab pH
IA1IL IL2KS KS2NE OK1WI 1:1 lineReg
4
5
6
7
8
9
4 5 6 7 8 9 Unprocessed MSP pH
Lab pH
IA1IL IL2KS KS2NE OK1WI 1:1 lineReg 4
5
6
7
8
9
4 5 6 7 8 9 Universal MSP pH
Lab pH
IA1IL IL2KS KS2NE OK1WI 1:1 lineReg
R^2 = 0.47 R^2 = 0.
R^2 = 0.60 R^2 = 0.
R^2 = 0.81 R^2 = 0.
2.5 Acre Grid
Raw MSP Universal MSP
Adjusted MSP Shifted MSP
Field Average
Numeric AgroNumeric Agro--Economic ModeEconomic Mode
NRCL = f (Income, Cost)
Income = f (Soil pH) Soil pH = f (True pH, Lime, Probability)
Cost = f (Lime)
Lime = f (Estimated pH)
Estimated pH = f (True pH, Probability)
True pH = f (Probability)
Model Input Modules:
- Categorized distributions
- Categorized functional relationships
- Multidimensional arrays
Model Output (Net Return over Cost of Lime)
Ys CL Q L d
Ps Yc d
Pc Ys d
Ps Yc d
Pc NRCL − ⋅
= (^1 ) 1 ( 1 ) ( 1 ) ( 1 )
d = annual discount rate Yc = corn yield Ys = soybean yield Q (^) L = prescribed lime application rate
NRCL = Net return over cost of lime Pc = price of corn Ps = price of soybean CL = cost of lime
Integrated Agitated Soil Measurement Integrated Agitated Soil Measurement
1:1 solutions with 15 Nebraska soils
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7. Reference pH
Measured pH R 2 = 0.87 (0.91 means) RMSE (Precision) = 0.15 pH Reg. SE (Accuracy) = 0.20 pH
pH
1
10
100
1 10 100 Reference nitrate-nitrogen (CR), mg/kg
Measured nitrate-nitrogen, mg/kg
(^) R 2 = 0.32 (0.40 means) RMSE (Precision) = 0.17 pNO 3 Reg. SE (Accuracy) = 0.22 pNO 3
1 pNO^3
10
100
10 100 1000 Reference soluble potassium (AAS), mg/kg
Measured soluble potassium, mg/kg
R 2 = 0.54 (0.63 means) RMSE (Precision) = 0.10 pK Reg. SE (Accuracy) = 0.13 pK
pK
Integrated Agitated Soil MeasurementIntegrated Agitated Soil Measurement
4.0 4.5 5.0 5.5 6.0 6.5 7. Reference pH
Measured pH
R2 = 0.98 (0.99 means) RMSE (Precision) = 0.08 pH Reg. SE (Accuracy) = 0.09 pH
pH
10
100
1000
10 100 1000 Reference soluble potassium (AAS), mg/kg
Measured soluble potassium, mg/kg
R 2 = 0.95 (0.98 means) RMSE (Precision) = 0.05 pK Reg. SE (Accuracy) = 0.03 pK
pK
Reference tests:
- Soil pH
- glass ion-selective electrode
- RMSE = 0.05 pH
- Soluble potassium
- atomic adsorption spectroscopy (AAS)
- RMSE = 0.01 pK
- Residual nitrate
- cadmium reduction (CR)
- RMSE = 0.02 pNO 3
1
10
100
1 10 100 Reference nitrate-nitrogen (CR), mg/kg
Measured nitrate-nitrogen, mg/kg
R 2 = 0.48 (0.67 means) RMSE (Precision) = 0.13 pNO 3 Reg. SE (Accuracy) = 0.10 pNO 3
pNO 3
VRT Prescription VRT Prescription
6.0 6.5 7.0 7. Measured buffer pH
Measured soil pH and predicted buffer pH
Predicted buffer pHMeasured soil pH 1:1 line( )
LR = f (buffer pH)
Buffer pH = f (soil pH, CEC) R 2 = 0.
R 2 = 0.
0
200
400
600
0 200 400 600 Measured exchangeble K, mg kg -
Measured soluble K and predicted
exchangeble K, mg kg
Predicted exchangable KMeasured soluble K 1:1 line
K rate = f (exchangeable K)
Exchangeable K = f (soluble K, CEC)
R 2 = 0.
R 2 = 0.
Integrated Multiple Data LayersIntegrated Multiple Data Layers
Maps produced by Veris Technologies, Inc. (Salina, Kansas)
Soil pH & Clay & OM
= Lime Recommendation
Applicability of OnApplicability of On--thethe--Go Soil SensorsGo Soil Sensors
Residual nitrate (total nitrogen) Some Some OK
CEC (other buffer indicators) OK OK
Other nutrients (potassium) Some OK
Soil pH Some Good
Depth variability (hard pan) Some OK Some
Soil compaction (bulk density) Good Some
Soil salinity (sodium) OK Some
Soil water (moisture) Good Good
Soil organic matter or total carbon Some Good
Soil texture (clay, silt and sand) Good OK Some
Soil property H+ H+ (^) H+H
H+
Status of Implementation Status of Implementation
- Commercial
- Electrical conductivity
- Topography
- Soil pH
- Visual/near-infrared spectroscopy
- Available solutions
- Implement draft
- Ground penetrating radar
- Magnetic field
- Gamma-radiometry
- Upcoming solutions
- Capacitance (moisture)
- Residual nitrate and soluble potassium
- Soil mechanical resistance
- Machine vision
- Small scale topography
Sensor fusion
Directed (Guided) Sampling Directed (Guided) Sampling
- Directed sampling should be used to
calibrate and/or validate sensor data
- Directed samples should be collected
from relatively homogeneous field areas
away from the boundary and other
transitional areas
- Directed samples should cover the entire
range of sensor-based measurements,
especially toward low and high ends
- Directed samples should be physically
spread across the entire field
- It should be possible to process multiple
sensor-based data layers
Currently Considered Criteria Currently Considered Criteria
Homogeneity
Neighborhood variability
Even data spread
D-optimality
Even field coverage
S-optimality
Example of Objective FunctionExample of Objective Function
OF = Sopt ⋅ Dopt − pH ⋅ Dopt − EC ⋅ Hcr − pH ⋅ Hcr − EC
- S-optimality
- D-optimality (soil pH)
- D-optimality (soil EC)
- H-criteria (soil pH)
- H-criteria (soil EC)
Prescribed SamplingPrescribed Sampling
4.5 5.0 5.5 6.0 6.5 7.0 7.5 8. Soil pH
Soil EC, mS/m
pH - L EC - H
pH - L EC - L
pH - H EC - L
pH - H EC - H
4.5 5.0 5.5 6.0 6.5 7.0 7.5 8. Soil pH
Soil EC, mS/m
a) b)
Categorical data
separation
Latin Hypercube
Sampling (LHS)
Summary Summary
- On-the-go soil sensors can provide high
density information about soil properties
- Many sensor approaches are past initial
commercialization stage
- Sensor fusion provides the ability to separate
various agronomic effects
- Site-specific sensor calibration and validation
are essential steps of the mapping process
- Laboratory soil analysis remains a required
supplementary practice
- Agro-economic value of selected sensor-
based data layers is site-specific http://bse.unl.edu/adamchuk
E:mail: vadamchuk2@.unl.edu
