Figure 1. Real-time (left) and map-based (right)
approach to use vehicle-based on-the-go soil sensors.
On-the-Go Soil Sensors for Precision Agriculture
A summary of real-time sensors and their applications in agriculture.
When looking at a yield map, one can observe how variable is crop performance in a field. However, the yield map does not give much information about the reasons for differences in crop yield. In some cases, it is related to special management practices or past history. However, frequently we can relate low yield to crop stresses associated with non-optimal soil conditions. Soil maps help us better understand the spatial patterns in soil properties that cause yield variability. A proper decision-making process prior to variable rate application must take into account soil variability. As a result, higher profits from implementation of precision agriculture are expected while reduced waste will help in maintaining the quality of our environment.
Currently, soil maps representing various properties are commonly obtained through recommended soil sampling and analysis procedures. Geo-referenced soil sampling, laboratory analysis, and mapping are available through several commercial vendors. The resulting interpolated soil maps become key information layers in prescribing variable rate application of fertilizers, lime and herbicides. Conventional soil sampling and analysis have shown mixed economical returns due to the high costs associated with labor-intensive sampling and analysis procedures and map uncertainties. In many cases, when the sampling density was not large enough, the limited number of soil samples did not produce an accurate representation of soil properties (especially for nutrient levels).
Several researchers as well as commercial institutions have been investigating various measurement techniques that could be suited to automated soil mapping (similar to yield monitoring). The sensors developed could be used either to control variable rate application equipment in real-time (Figure 1 left) or in conjunction with GPS to generate field maps of particular soil properties (Figure 1 right). Depending on the spacing between passes, travel speed, and sampling and/or measurement frequency, the number of points of measurements per acre varies; but in most cases, it is much greater than the density of manual grid sampling. The cost of mapping could be reduced as well. The purpose of this publication is to review the most promising soil sensor approaches and to present an overview of some that are currently commercially available.
|
|
|
Figure 1. Real-time (left) and map-based (right)
approach to use vehicle-based on-the-go soil sensors.
|
When thinking about an ideal precision agriculture system, producers visualize a sensor located in direct contact with, or close to, the ground and connected to a “black box”, which analyzes sensor response, processes the data, and changes the application rate instantaneously. They also hope that the real-time information detected by the sensor and used to prescribe the application rate would optimize the overall economic or agronomic effect of the production input. This approach, however, does not take into account several difficulties that are seen in the “real world”:
Rather than using real-time, on-the-go sensors with controllers, a map-based approach may be more desirable because of the ability to collect and analyze data, make the prescription, and conduct the variable rate application in two or more different steps. In this case, multiple layers of information including yield maps, a digital elevation model (DEM), and various types of imagery could be pooled together using a geographic information system (GIS) software package designed to manage and process spatial data. Prescription maps can be developed using algorithms that involve several data sources as well as personal experience. Probably the most essential piece of data is a set of maps representing variation in soil characteristics that influence yield, such as:
Sensors for Automated Measurements
Scientists and equipment manufacturers are trying to modify existing laboratory
methods and develop indirect measurement techniques that could allow on-the-go
soil mapping. To date, only a few types of sensors have been investigated, including:
Electromagnetic sensors use electric circuits to measure the capability for soil particles to conduct or accumulate electrical charge. When using these sensors, the soil becomes part of an electromagnetic circuit, and changing local conditions immediately affect the signal recorded by a data logger. Several such sensors have become commercially available. For example, one way to estimate soil electrical conductivity (EC) is by electromagnetic induction using a commercially available Geonics Limited EM38* meter. The transmitting coil induces a magnetic field that varies in strength with soil depth. The magnetic field strength/depth to soil relationship can be altered to measure different depths of the soil to a maximum depth of 1.5 meters. A receiving coil measures the primary and secondary “induced” currents in the soil and relates the two to soil electrical conductivity. Another commercially available instrument for mapping soil EC, Veris® EC Probe, measures EC more directly. It uses a set of coulter electrodes that send out an electrical signal through the soil. The signal is received by two sets of electrode coulters that measure voltage drop due to the resistivity of the soil, indicating soil EC for two different depth ranges. (See detailed article on soil electrical conductivity).
Optical sensors use light reflectance to characterize soil. These sensors can simulate the human eye when looking at soil as well as measure near-infrared, mid-infrared, or polarized light reflectance. Vehicle-based optical sensors use the same principle technique as remote sensing. Several researchers have worked on the development of optical sensors to predict clay, organic matter, and moisture content. Rather than using optical reflectance, some researchers are utilizing ground-penetrating radars to investigate wave movement through the soil. Changes in wave reflections may indicate changes in soil density or restricting soil layers. (See detailed article on remote sensors.)
Mechanical sensors can be used to estimate soil mechanical resistance
(often related to compaction). These sensors use a mechanism that penetrates
or cuts through the soil, and records the force measured by strain gauges or
load cells. Several researchers have developed prototypes that show the feasibility
of continuous mapping of soil resistance, however, none of these devices is
commercially available. The draft sensors or “traction control” system
on tractors use a similar technology to control the 3-point hitch on the go.
Electrochemical sensors could provide the most important type of information
needed for precision agriculture – soil nutrient levels and pH. When soil
samples are sent to a soil-testing laboratory, a set of standardized laboratory
procedures are performed. These procedures involve sample preparation and measurement.
Some measurements (especially determination of pH) are performed using an ion-selective
electrode (with glass or polymer membrane, or ion sensitive field effect transistor).
These electrodes detect the activity of specific ions (nitrate, potassium, or
hydrogen in case of pH). Several researchers are trying to adapt existing soil
preparation and measurement procedures to essentially conduct a laboratory test
on the go. The values obtained may not be as accurate as a laboratory test,
but the high sampling density may increase the overall accuracy of the resulting
soil nutrient or pH maps.
Airflow sensors
were used to measure soil air permeability on-the-go. The pressure required
to squeeze a given volume of air into the soil at fixed depth was compared to
several soil properties. Experiments showed potential for distinguishing between
various soil types, moisture levels, and soil structure/compaction.
Acoustic sensors have been investigated to determine soil texture
by measuring the change in noise level due to the interaction of a tool with
soil particles. Low signal-to-noise ratio did not allow this technology to develop.
Sensor Data Usage
Although various vehicle-based soil sensors are under development, only electromagnetic
sensors have been commercialized and widely used to date. Ideally, producers
would like to operate sensors that provide direct inputs for existing prescription
algorithms. Instead, commercially available sensors provide measurements such
as electrical conductivity (EC) that cannot be used directly since the absolute
value depends on a number of physical and chemical soil properties such as:
texture, organic matter, salinity, moisture content, etc. Alternatively, electromagnetic
sensors give valuable information about soil differences and similarities that
make it possible to divide the field into smaller and relatively consistent
areas referred to as management zones.
For example, such zones could be defined according to various
soil types found across a field. In fact, soil EC maps usually can better reveal
boundaries of certain soil types than soil survey maps (used for rural property
tax assessment). Different anomalies such as eroded hillsides or ponding can
also be easily identified on a soil EC map. Figure 2 compares a soil survey
and a soil EC map for the same field showing some differences in boundaries.
 |
|
Figure 2. Soil EC map (right) compared to soil survey
(left).
|
Crop yield maps also frequently correlate to soil EC maps. In many instances such similarities can be explained through differences in soil. In general, the soil EC maps may indicate areas where further exploration is needed to explain yield differences. Both yield potential and nutrient availability maps may have a similar pattern as soil texture and/or organic matter content maps. These patterns can often be revealed through a soil EC map as well. Therefore, it seems reasonable to use on-the-go mapping of electromagnetic soil properties as one layer of data to discover the heterogeneity (differences) of soil within a field (similar to using bare soil imagery). Zones with similar electrical conductivity and a relatively stable yield may receive a uniform treatment that can be prescribed based on a reduced number of soil samples located within the zones on the EC map.
As new on-the-go soil sensors are developed, different real-time and map-based variable rate soil treatments may be economically applied to much smaller field areas that could reduce the effect of soil variability within each management zone.
Summary
More accurate soil property maps are needed to succeed in implementing site-specific
management decisions. Inadequate sampling density and the high cost of conventional
soil sampling and analysis have been limiting factors. On-the-go, vehicle-based
soil sensors represent an alternative that could both improve the quality and
reduce the cost of soil maps. When further developed, on-the-go soil sensors
may be used for either real-time or map-based control of agricultural inputs.
To date, only the mapping of electromagnetic soil properties is available commercially.
These maps can be used to define management zones reflecting obvious trends
in soil properties. Each such zone can be sampled and treated independently.
Smaller management zones will be feasible when new on-the-go soil sensors are
developed and commercialized.
*Mention of brand names is for identification purposes only. No endorsement or criticism intended for those mentioned or any other equivalent products.
