National GIS Manager
Airesearch Mapping Pty Ltd.
11 Hi-Tech Ct, Technology Park
Eight Mile Plains QLD 4113
Ph. 07 3841 3433 Fax. 07 38413466
Using a Geographic Information System (GIS), analysis of an entire farm can be made down to sub-paddock regions. Information such as soil, slope, aspect and rainfall can all be analysed to develop a picture of the characteristics of each parcel of land. Characteristics of one parcel can be compared to others to determine why yields and costs incurred vary from one location to another.
With the combination of remotely sensed data (satellite imagery, aerial photography, plus others) and GIS, decisions on farm management can be made to maximise returns and reduce costs.
Imagery can be used to determine the health and vigor of crops and assist in determining growth rates. Informed decisions can then be made on the productivity of various parcels. Action can be taken to improve areas that are not performing.
Yield returns can be calculated and risk assessments made for the viability of any parcel of land. Long-term strategies can be formed with confidence that the information being used is the best available. With the information determined from imagery the harvest could be planned to maximise yields over the entire farm.
This paper will identify the variety of airborne and satellite sensors available to provide what is one of the essential and fundamental data sets to this equation – imagery, and discuss a broad range of applications for it’s use.
Imagine the scenario that it is only a few weeks after planting your fields. You get into your tractor and drive it out to one of your fields where you survey the fledging crop in front of you. You reach out and turn the Global Positional System (GPS) receiver on and your location is displayed possibly down to a metre. You reach over and press another button and a series of maps are displayed onto a screen.
These maps display characteristics of your field such as, soil type, topsoil depth, slope, soil nutrients, etc. Next, you download some recent imagery that indicates where your crop is performing and where it is struggling. The machine then uses this data to regulate the application of fertilisers over the crop in the field.
While this may seem like some futuristic or far-fetched scenario, it is actually happening at this very moment. While this is the ultimate use of all of the current technology it is not beyond the means of most farmers to begin using some of this information. You can start using the technology in a less substantial way and still gain practical benefits.
The advantage of knowing your precise location means that many variables can be recorded and logged against specific areas. Fertilizers, insecticides and seeding rates can be regulated to fit the properties and condition of the soils. Yield data can be recorded during harvesting and future management of the area can be made relying on scientific data instead of hunches.
Geographic Information Systems (GIS)
A computerised system of hardware and software that is designed to store, retrieve, display, manipulate and analyse spatial or geographic data.
Global Positioning Systems (GPS)
A network of 24 U.S. military satellites that allow ground-based receiver units to determine their location.
The detection of objects with a sensor that is not in direct contact with the object. The recording of reflected light or wavelengths produced by different objects on the earth’s surface e.g. A photograph taken with a typical hand-held camera.
GIS refers to computer hardware and software that manages geographical based data. The system is able to store, retrieve, transform, query and manipulate data. GIS software for precision farm management will store data for field parameters (aspect, slope, etc), soil properties (chemical, physical, biological) and images into discrete layers. This layer information will be recorded against the particular field location.
Recording the relationship between variable attributes in the same spatial location will allow for sound queries to be formulated. A fully functional GIS will be able to query and display the data in a variety of ways that is useful to the end user.
Characteristics can be analysed to display application maps, yield maps or other required options. The information in a GIS needs to be stored in a coordinate system for the layers to be able to relate to each other. Once the coordinate system is established characteristics can be analysed and displayed. Maps can be created for application rates or yield totals plus a myriad of other features. These maps may be displayed electronically or printed as hard copy products.
GIS can aid in a better understanding of interactions between yield, fertility, pests, weeds and other crop related factors.
Variable rate technology uses computers to control the application of chemicals and fertilizers. There are two ways these controllers work.
The first is via an application map and GPS to locate your field position. The GPS tells the tractor where its location within the field is. The GIS uses this positional information from the GPS to access data about the field at that location. Information is then sent to the controller about the field conditions. Using predetermined calculations the controller will allow the required amount of applicator to be distributed.
The second method is based on sensors within the tractor, which record field data as it travels along. As the data is received the tractors controller regulates the rates of dispersal of either the fertilizer or insecticide.
Other developments in this field include VRT irrigation systems and VRT seeding applicators.
Remote sensing works by capturing light reflected from the earth’s surface with a camera or sensor system usually mounted on an aerial or space borne platform. Natural and man-made features absorb, transmit and reflect different wavelengths of light in varying quantities. Sunlight interacts with materials such as soil and plants to gather information on these features, which can be captured, via the sensors. Remote sensing in effect uses the same principles as the human eye.
Imagery is often referred to in terms of spatial and spectral resolution. The characteristics of these two factors will help determine the product that best suits the application.
The ability of the sensor to detect energy emitted or reflected by surface features within certain parts of the electromagnetic spectrum is known as spectral resolution. Spatial resolution on the other hand refers to the size of the objects that can be distinguished within an image. Images are made up of rectangular cells (pixels) that correspond to a particular distance on the ground. So a spatial resolution of 10 metres will not clearly detect an object that is less than 10 metres.
Remote sensing sensors can be designed to capture electromagnetic energy at various wavelengths from either end of the spectrum. A representation of the electromagnetic spectrum is shown in Figure 1 above 2. The visible to infrared range is where most agricultural remote sensing is performed. This range of wavelengths allows us to exploit the physical properties of natural materials such as plants and water.
Following is a list of some of the more popular and well-known types of remotely sensed data.
MSS - Multi Spectral Sensors
Multi spectral sensors record reflected or emitted energy in several bands within the electromagnetic spectrum. As an example, one set of detectors will record reflectance from visible blue energy while another will record from infrared energy. The capture of the multiple reflectances can be combined to make colour images. Current multi spectral systems measure between 3 to 7 bands simultaneously.
Panchromatic (pan) sensors record the energy reflectance in one wide band of the electromagnetic spectrum. The width and location within the spectrum may vary depending on the sensor involved. For most current pan sensors the single band usually spans the visible to near infrared area of the spectrum. Panchromatic data is always displayed as black and white.
SAR - Synthetic Aperture Radar
MSS and Pan sensors are both passive systems. That is they record energy reflected from another source such as the sun. SAR sensors are active systems that transmit signals in the microwave portion of the spectrum. The sensor records the strength and orientation of the return signal from the target. Because SAR is an active sensor and uses longer wavelengths than the optical sensors it is capable of acquiring images under virtually all weather conditions. They can record through clouds, fog and darkness.
Aerial photography usually captures in the visible to infrared part of the spectrum. Aerial photography has not changed much over the past few decades except for natural advances in technology for lens, film and processing. Aerial photography has an advantage over satellite imagery in the spatial resolution that can be achieved. If sub-metre resolution is required then aerial photography can deliver at reasonable costs.
The vegetation photosynthetic process is where remote sensing can play an important part in agricultural production. During photosynthesis plants grow and release oxygen into the atmosphere by converting sunlight, water and carbon dioxide.
The human eye is only sensitive to the portion of the spectrum called visible light. This light (blue, green, red) ranges from around 400 nanometres to 700 nanometres in wavelength. The part of the spectrum typically used for remote sensing in agriculture ranges from between 400 – 1500 nanometres. The range from 700 –1500 nanometres is the near infrared range, which is far beyond the human eye, can see. It is in the near Infrared (NIR) range of the spectrum that healthy growing plants react differently to sunlight.
Plants reflect and absorb sunlight. The absorbed sunlight is used during photosynthesis to create new material for the plant. Plants absorb more red and blue light in the visible part of the spectrum than they do green light. It is the absorbed red and blue light that fosters photosynthesis and the greater reflectance of green light that gives plants typically a green look to the human eye.
Infrared photography was originally used by the military to emphasise the difference in reflectance between live, healthy vegetation and green paint for camouflage, which portrays similar characteristics in the visible spectrum. Since then infrared has become particularly useful for a variety of crop survey projects.
Figure 2 -Reflectance of 3 objects through the visible and near infrared wave
The wavelengths of near infrared are too long for the plant to be able to use during photosynthesis. Therefore the plant absorbs very little of this light but instead reflects most of it back into the atmosphere.
Hutchins states infrared wavelengths have a far higher reflectance than any other wavelength from the structural material within plants, refer Error! Reference source not found.. Plants, which are under stress due to disease or other environmental problems, have a decreasing amount of photosynthetic activity. With this decrease in photosynthesis more visible light is reflected than absorbed. This photosynthetic behaviour has no influence on the infrared reflectance. However problems with crops related to poor development and health will affect both visible and infrared responses.
In the visible part of the spectrum the vegetation has a peak in the visible green and that is why we see vegetation as green.
- The vegetation curve has a dip in the visible part of the spectrum at the colour red. This is due to healthy vegetation absorbing light at this wavelength during photosynthesis.
- In the near infrared, healthy vegetation has a much higher reflectance than soil. Contrast this to the similarities in the visible spectrum. This is due to the structure of leaves having high reflectance in the near infrared.
Pictures of crops at a specific wavelength can be useful but are far more useful when combined with images from other wavelengths. Vegetation indicies are created by using ratios of different wavelengths to produce pictures or maps. The most common of all these indicies for providing an indication of crop productivity and regeneration is the NDVI (Normalised Difference Vegetation Index). The NDVI is calculated by the formula
Where NIR = Near Infrared & R = Red reflectance
Results from this formula range from +1 to –1.
Figure 3 - Results from an NDVI classification. The range is from blue (no vegetation) to darkest green (healthy vegetation).
In healthy crops there is a large difference between the reflectance of both parts of the spectrum. Red is largely absorbed while NIR is largely reflected. This difference transfers to a positive index.
Other vegetation indices exist by using other wavelengths in different ratios. The RVI (Ratio Vegetation Index) divides one wavelength by another. The SAVI (Soil Adjusted Vegetation Index), which resembles the NDVI with some adjustments, made for the brightness of background soil.
Problems within crops like moisture (too much or too little), insect infestation, disease, nutrient deficiencies and other stresses can be calculated by using these different wavelengths in different ratios.
Visual detection of plant stress is one method of using remote sensing. This method uses light in the visible spectrum only. By using a more scientific method of analysis and by using wavelengths outside the visible spectrum it is expected that better and more sensitive detection of plant-associated problems will be picked up before they become apparent to the naked eye.
Variations of crop yield within a parcel are possibly the most important input for crop management. Crop yield is related to a vast array of crop and soil characteristics such as moisture, nutrients, pestilence, slope and numerous others. Linking the spatial information on these two inputs, crop yield and field characteristics, through a GIS can allow for an analysis of the factors that are present in determining crop production. The resultant analysis then becomes the benchmark for strategies to improve crop yield.
The yield measurements can be analyzed to determine the effectiveness of VRT inputs. These inputs can be refined for future years production.
Figure 4 - The image on the left is a map of yield. The range from green to blue represents highest to no yield. The right image is a map of soil acidity. The range from green to red represents lowest to highest.
Applications of remote sensing in agriculture may only be limited by a persons imagination. Current applications are designed to provide the farmer with timely information about crop progress. The following are just some of the benefits that can be gained from the use of remote sensing.
- Early identification of crop health and stress
- Ability to use this information to do remediation work on the problem
- Improve crop yield
- Crop yield predictions
- Reduce costs
- Reduce environmental impact
- Crop management to maximise returns through the season
- Crop management to maximise returns during harvest time.
It can be seen that remote sensing data used appropriately and at the right times of the season, has the ability to provide benefits to crop health and hence improve production.
The precision farming approach to crop production may be viewed as a four step process (Sudduth) see Figure 5.
Figure 5 - The cyclic nature of the precision farming approach.
Using data obtained through GIS, GPS and imagery many functions can be planned from the start of the crop. Seeding rates across the field can be varied to allow for better plant growth. To allow crops to mature properly based on the lay and condition of the land seed spacing can be varied across a field. A digital map of the field and GPS can allow a tractor to vary seeding rates based on field characteristics like soil type, slope, and aspect. Seeding rates and hence yield rates can be maximised according to soil conditions.
The application of fertilizers and insecticides can be varied in much the same rate across the field as seeding rates. The dispersal of fertilizers and insecticides based on soil characteristics will provide positive outcomes economically and for the environment.
Harvesting equipment can be equipped with services that measure the yield as you go. Combined with GPS a yield map can be produced to show variations in yield. Low yield maps can be inspected and appropriate measures under taken for the next cropping season.
While the use of remote sensing in agriculture is still relatively new it is far from being in its infancy. Large numbers of farmers are already making use of its benefits and as time goes on and techniques improve it may be a brave (or foolish) farmer who doesn’t.
Accessed On Line
Rains G C, Thomas D L; http://www.ces.uga.edu/pubcd/B1186.htm, accessed 24/3/01
http://www.uswcl.ars.ag.gov/epd/remsen/rsagintr.htm, accessed 24/03/01
http://www.uswcl.ars.ag.gov/epd/remsen/Vi/VIworks.htm, accessed 24/03/01
Hutchins K S, http://www.imagingnotes.com/summer98/summer2.html, accessed 24/03/01
Walters A, Baumgart J, Knapp J; http://www.imagingnotes.com/summer98/summer3.html, accessed 24/03/01
Kelly R, Jensen T, Harris G; http://www2.dpi.qld.gov.au/dpinotes/fieldcrops/genmanage/fs00405.html, accessed 24/03/01
Peterson T; http://www.pioneer.com/usa/technology/remotesum.htm, accessed 15/04/01
Sudduth K A; http://www.fse.missouri.edu/mpac/pubs/eng_tech.pdf, accessed 15/04/01
Yang C, Anderson G L, Everitt J H; http://www.precisionfarming.com/features/0199pf_yang.html, accessed 8/05/01
Clark R L; http://www.precisionfarming.com/features/0497Clark.html, accessed 8/05/01
Srinivasan A; http://www.vtt.co.jp/staff/ancha/pfasia.html, accessed 15/04/01