Remote Sensing in Agriculture using Radiometers
Using remote sensing to quantify and
control nitrogen stress in corn.
Site specific management or precision farming has been defined as applying the right amount of agricultural input (water, nitrogen, herbicide, etc.) in the right place at the right time. Remote sensing using radiometers has potential as a data collection technique to assist management decisions for in-season water and nitrogen management.
|
| Figure 1. The electromagnetic spectrum. |
Remote sensing is the measurement of some property of an object
by a recording device that is not in physical contact with the object under
study. Reflected or emitted radiation in the electromagnetic spectrum is the
source of energy capable of conveying information about the object of interest.
Figure 1 represents the ordered array of known electromagnetic
radiation that extends from the very short wavelengths of gamma rays to the
long wavelengths of radio energy. Within this continuum of energy, the human
eye is only sensitive to the visible region which ranges from blue light (4
x 10-5 cm or 400 nm) to red light (7 x 10-5 cm or 700 nm). The visible, infrared
(IR), and the microwave regions of the electromagnetic spectrum are of primary
interest to agriculture. This article focuses on the visible and near infrared
(NIR) region for detecting information within a soil/crop scene. Figure 2 shows
spectral curves for bare soil and mature corn in the visible and NIR regions
of the electromagnetic spectrum. The energy reflected from a bare soil surface
is much different than energy reflected from a mature, healthy corn crop when
very little soil can be seen through the plant canopy. Bare soil reflectance
is essentially a straight line which increases as wavelength increases. A healthy,
green plant strongly absorbs blue and red light due to the presence of chlorophyll
in the leaves, moderately reflects green light because of its color, and strongly
reflects NIR light due to scattering within the canopy. These particular properties
are known as spectral characteristics. They indicate where the most information
can be obtained by concentrating on these areas of the spectrum. Thus, instead
of using expensive spectroradiometers which produce spectral curves as shown
in Figure 2, radiometers that measure radiation in predetermined areas of the
electromagnetic spectrum are used. This provides a more economical approach
to data collection.
 |
| Figure 2. Spectral reflectance of bare soil and a mature corn canopy. |
Radiometers are nonimaging instruments in the sense that they do not produce a picture. These instruments integrate over the area within their field-of-view to produce a single number that characterizes the electromagnetic energy emitted or reflected by the object. A radiometer typically has three major components; these are (1) the optical system which consists of lenses and filters, (2) the detector which provides an electrical signal proportional to radiant energy impinging on its active surface, and (3) the signal processor which performs specified functions on the electrical signal to provide the desired output data. The four channel radiometer shown in Figure 3 has filters that only allow blue light (450-520 nm), green light (520-600 nm), red light (630-690 nm), and NIR light (760-900 nm) to strike the detector in each respective channel. These wavebands are similar to the four spectral bands used in the ThematicMapper onboard the Landsat series of satellites, the Ikonos satellite operated by SpaceImaging, and the QuickBird satellite operated by DigitalGlobe.
Radiometers have been manufactured with as few as two broad band channels and as many as 16 narrow band channels. Virtually any spectral band in the 400 to 1100 nm range with a bandwidth of 10 nm or more can be used in these instruments. However, narrow bandwidth filters decrease the amount of light that strikes the detector which means that additional electronics must be incorporated into the instrument to obtain useful output signals.
 |
|
| Figure 3. Four channel radiometer manufactured by Exotech Incorporated. | |
Radiometers can be used in various ways from handheld operation for detailed studies on individual plant leaves to mounting in aircraft to obtain spectral data representative of large areas. Figure 4 shows radiometers attached to an extendable boom which is mounted on a high-clearance tractor for field access. The downward looking radiometer on the end of the boom measures energy reflected from the crop while the upward looking radiometer on the tractor’s roll over protection system simultaneously measures incoming radiation. A circular area 2.5 m in diameter is viewed by the down-looking radiometer from a height of 10 m; approximately 30 corn plants are contained within this area. Data from the radiometers are recorded every two seconds as the tractor moves through the field. GPS coordinates are also taken with each data point to determine location of the sensor in the field. Data transects through the field are taken at 24-row intervals. Crop reflectance in each of the four spectral wavebands is calculated as a ratio of the energy measured from the crop divided by the incoming energy.
An example for use of radiometer data to assist with in-season nitrogen management on irrigated corn to improve nitrogen use efficiency is presented below. Corn leaves with a nitrogen deficiency contain relatively little chlorophyll and tend to have a pale green color; thus, reflectance in the green portion of the visible spectrum increases as plant nitrogen deficiency increases. Changes in the red area of the spectrum show small increases in reflectance, but not as abrupt as the green reflectance. Corn canopy reflectance in the NIR portion of the electromagnetic spectrum tends to decrease as plant nitrogen deficiencies increase. Therefore, the green and NIR reflectance values can be used to calculate the Nitrogen Reflectance Index (NRI) to determine when and where in a field nitrogen should be applied to keep crop growth and yield at an optimum level without excess applications of nitrogen. Figure 5 is an example of a nitrogen sufficiency map for irrigated corn at its 8th leaf growth stage generated from NRI information. Light colored areas within the field had NRI values less than 0.95 which indicate early signs of nitrogen deficiency. The dark colored areas indicate that the corn is healthy and not in need of additional nitrogen at that particular time. Applying nitrogen to the light colored areas would be the recommended practice at this particular growth stage to reduce the amount of nitrogen applied to the field and reduce input costs.
|
|
||||||||||||
|
Walter Bausch |
||||||||||||
