Question: Why does AMS make such a big deal about using "far-IR" (far infrared) imagery? It is just another form of light, isn't it?

Answer: Yes, all light comes in the form of waves. It is all the same except for the length of the waves; i.e., the "wavelength" of each form of light. Light that is visible to the human eye ranges from violet on the shorter wavelength side to red light on the longer wavelength side, with the wavelengths of all the various colors falling in between.

Most light is invisible to the human eye. Just slightly longer than red light is an invisible form called "near infrared". Much longer than near infrared is a form that, if sufficiently intense, would be felt on your hand as heat. Such light is called "far infrared" and this region is referred to as the "thermal region" of the spectrum. (Please note that these two regions, the near infrared and the far infrared, are totally distinct. It's important not to confuse one with the other.)

Wavelengths of visible light and near infrared are measured in ten-millionths of an inch. Far infrared wavelengths are several tenths of an inch. Waves longer than far infrared are microwaves, about one-third of an inch, and radio waves are measured in feet to hundreds of feet. Shorter than violet light, just beyond the short end of the visible spectrum, is "ultraviolet," followed by "x-rays" and "gamma rays."

The four kinds of light used for agricultural remote sensing are: visible, near infrared, far infrared and (on an experimental basis) ultraviolet. Most remote sensing systems used in agriculture utilize visible and near infrared light because knowledge of how to build such systems is widespread and such systems are relatively easy to build.

The AMS system operates in the far infrared (i.e., the "far-IR") region of the spectrum. The reason AMS choose far infrared is that, according to the results of NASA's Corn Blight Watch Experiment, most of the information concerning agricultural problems (i.e., stress) is contained in the far infrared region. A small amounted of such information is also contained in the near infrared and very little - except for advanced stage problems - is accessible in the visible region of the spectrum.

Another advantage of a properly designed thermal system is sensitivity. This means that, other things being equal, one can detect problems earlier when they have a better chance of being treated successfully. In addition, as noted elsewhere, AMS has introduced the decisive advantage for agricultural applications of "clutterless" imagery.
Back to Top

Question: What are the different types of agricultural remote sensing, what is sensed by them, and how does remote sensing work?

Answer: Visible spectrum imagery consists of photographs, taken either from an aircraft or from a satellite. As the name implies, these types of images are what the normal human eye sees. Its advantage is that it is familiar and provides a photographic image that can be studied, enlarged, etc. Certain land features can be detected from the air with such imagery. Its use in assessing crop stress is limited to showing damage that is already visible. While of use in directing attention to remote or hard-to-reach areas (such as treetops, forests), it is of very limited use in agriculture that produces fruits and vegetables, since visible stress indicates only already-present damage and cosmetic appearance is very important in those crops.

Near-IR (near infrared) or CIR (color infrared) photography consists of use of film that is sensitive to visible and near infrared light, exposed in a camera flown in an aircraft. It is not heat-sensitive because it is not sensitive to the far infrared portion of the spectrum. It is the most common form of agricultural remote sensing at present; it is low cost and low tech. Like visible spectrum imagery, CIR photography will only show advanced-stage, high-stress problems. It suffers badly from false positives (apparent problems that are due to atmospheric clutter caused by clouds and other conditions common during the day) and masking of agricultural problems and conditions that fail to show up in the photograph. The stress levels it displays are not accurate, and it displays them as a large number of different shades of red. Dead vegetation shows up as blue.

The charge-coupled device (CCD) camera is the second most commonly used remote sensing system in agriculture. It usually has three sensors. A single CCD sensor consists of many thousands of mini-sensors lined up in rows and columns on a chip. Each of these mini-sensors "looks" at a different spot (pixel) in the field when an image of the field is captured. Any chip containing these thousands of mini-sensors is called a CCD array. Most CCD cameras have three CCD arrays: one is covered by a green photographic filter that lets through only green light; another uses a red filter; the third a near infrared filter. Thus, three separate images of the field are electronically recorded. The green and red images are often combined into a single image. Sometimes all three images are combined to produce an electronically generated version of a color infrared photograph.

CCD cameras are more sensitive than CIR photography but not nearly as sensitive as a properly designed thermal infrared system. CCD cameras suffer from masking and from false positives. CCD cameras are found in both aircraft and satellites.

CCD arrays have to be calibrated frequently since the individual sensors in the array are far from uniform and constantly changing. In satellites this is not possible after launch: the mini-sensors continue to change, and this eventually degrades the sensor's imaging ability. CCD cameras in aircraft have to be removed and recalibrated frequently or the quality of their imagery decays over time. The "Thermal IR Scanner" is the least used instrument in remote sensing, perhaps because it is more complex mechanically and is the only system with moving parts. The scanner focuses the infrared signal onto a detector and outputs an electronic signal, which is proportional to the amount of infrared signal it receives. This signal is stored in memory and the computer constructs an image from this signal. Thermal scanners, of the sort used by the AMS system, are those that image the crop with far infrared light. AMS uses scanners that detect the crop in the far-infrared (thermal) area of the spectrum. AMS scanners are seven times as sensitive as typical remote sensing systems used in agriculture. This enhanced sensitivity is the result of cooling the sensor to liquid nitrogen temperature (about 350°F below zero) and certain other proprietary techniques that significantly enhance sensitivity.

The AMS scanner, cooled to liquid nitrogen temperature, is sensitive to changes in a plants temperature and "emissivity". Emissivity changes when the water or chlorophyll content of plants change. The AMS system can thus detect changes in temperature, chlorophyll, and water content of plants from one field pixel to another. And these are the parameters of the plant that change when plants are stressed. In a stressed plant, temperature rises, water content may decrease, and chlorophyll content may decrease. Only far-IR permits early detection of changes in all these parameters.
Back to Top

Question: Why not put far-IR sensors on satellites?

Answer: Thermal far-IR scanners have the disadvantage of rapidly moving parts, making them mechanically complex relative to other remote sensing systems. Such parts don't last long in space, making scanners impractical for use aboard satellites. Also, there is a tradeoff between temperature sensitivity and pixel size. A temperature sensitivity of 2 degrees Centigrade would require a field pixel as large as a baseball diamond. Furthermore, if one tried to increase temperature sensitivity the pixel size would get even larger.

In an aircraft at several thousand feet a pixel size of two meters - six-and-one-half feet - is compatible with a temperature sensitivity of several tenths of a degree Centigrade. Thus, airborne scanners are appropriate for crop management of individual fields while a satellite based thermal sensor is not.

Finally, there is no practical way to cool sensors aboard a commercial agricultural satellite to liquid nitrogen temperatures.
Back to Top

Question: If far-IR imagery is so superior for detecting plant stress, why hasn't it been used more widely?

Answer:Scanner technology is primarily military and optimum designs are not widely known. Thermal scanners are inherently more complex to construct than remote sensors in the visible and near infrared regions of the spectrum; this is probably the main reason that the latter sensors are commonplace even though their performance in agriculture is less than satisfactory. In addition AMS has introduced the decisive advantage for agricultural applications of "clutterless" imagery.
Back to Top

Question: Aren't there other companies gathering imagery from aircraft?

Answer: Yes. Most other companies use visible, near-IR, or a combination of visible and near-IR sensing to produce their imagery. Only AMS imagery consistently detects the full range of plant stress, often before it becomes visible to the eye, without introducing false positives.
Back to Top

Question: How is it possible to detect stress before it is visible?

Answer: Plants initially under stress begin to experience small increases in temperature, and/or slight losses of chlorophyll and water. AMS's remote sensing system detects the temperature increase directly with great sensitivity and the loss of chlorophyll and/or water indirectly, by their effect on plant emissivity to which the system also responds.
Back to Top

Question: AMS says its imagery is "geo-referenced," using global positioning satellites. What does this mean? I have a GPS unit that I bought commercially. Why can't I use that with AMS imagery?

Answer: A suitably designed Geographical Positioning System (GPS) unit in the AMS aircraft and suitable AMS software allows the latitude and longitude of any point in AMS imagery to be located to within two meters (six and one-half feet). By clicking on any point in the imagery when it is displayed on a screen, one gets a read-out of latitude and longitude of that point that is accurate to within 2 meters. This allows areas of stress in the field - even pre-visible stresses - to be accurately located, provided the fieldman uses a suitably accurate GPS. AMS has found that most commercial GPS units used in agriculture can sometimes be off by as much as 100 feet. This is the reason that AMS has designed its own light backpack GPS units that are accurate to within two meters one hundred percent of the time. The ability to locate even small areas of pre-visible stress allow more accurate scouting, early and pre-visible detection and identification of field problems, definition of where to sample, and potentially, the replacement of grid sampling by AMS imagery-directed sampling.
Back to Top

Question: Suppose a fieldman sees nothing (because the imaging is picking up T stresses) when he gets out to a field location that the imagery shows as stressed. What does he do then?

Answer: If the fieldman is using conventional imagery he wouldn't be able to distinguish between a pre-visible stress at that location and a false positive displayed by the imagery. If the fieldman is using AMS imagery he knows it is a pre-visible stress because AMS imagery has been demonstrated not to display false positives. Being able to rely on the belief that it is pre-visible stress provides economic justification for the fieldman to take plant and soil samples from that location back to the lab to identify the nature of the stress.
Back to Top