Red River Growers Turn to Satellite to Manage Nitrogen
By: Kevin Corbley

Nitrogen management is a delicate balancing act in the sugar beet fields of the Red River Valley bordering North Dakota and Minnesota. Growers there tread a fine line between total production of their harvest and overall quality of the individual beets. Too little nitrogen in the soil inhibits growth and reduces the tonnage of crop, but an overabundance causes the beets to grow too large, depleting their valuable sugar content in the process.
    "The objective is to produce the most sugar beet tons with the highest percentage of sugar in those tons," said Mike Kasowski, a crop consultant with Centrol Inc. of Twin Valley, Minnesota.
    Achieving that goal, however, is easier said than done. Centrol estimates that in any given year beet farmers are battling nitrogen variability problems in 75% of their fields. Most of the time, the problem is caused by variations in nitrogen levels within a field, according to Kasowski.
    The traditional approach to preparing a field for sugar beets has been to apply a uniform dose of nitrogen across the entire field so that the residual nitrogen in the soil plus the amount applied equals about 120 pounds per acre in the Red River Valley. Difficulties arise when the fertilizer is spread on fields, or parts of fields, where excess residual nitrogen already lies in the soil.
    "We pay for too much nitrogen in the soil in terms of poor quality in the next [sugar beet] crop," said Darrell Larson, a grower in Trail County, North Dakota.
    Larson points out that too much nitrogen can harm other crops as well in the valley where small grains are often planted for two or three years between beet crops. Since the fertilizer is often applied immediately after beet harvest, an excess can cause wheat to lodge or have higher disease potential.
    Soil sampling has had only limited success in determining residual nitrogen content in beet fields for two reasons. First, taking one composite sample from an 80-acre field before sugar beets cannot detect variability across that field. And second, much of the residual nitrogen left behind after harvest in a sugar beet field is found in the plant foliage, not in the soil.
    Centrol, a farm management cooperative that has been working with growers in the Red River Valley since the early 1980s, is using a combination of precision agriculture technologies - satellite imagery, grid sampling and variable rate spreading - to prescribe just the right dose of nitrogen to maximize profits for beets and crops following them in the valley.

Satellites Detect Variability
Satellite imagery has been used for several years to alert growers to the presence of small pockets of crop variability within a field so they can investigate the cause and hopefully remedy the problem before damage is done. Variability is also the key to Centrol's nitrogen management project, but in this case variations in the field are correlated to residual nitrogen content after harvest.
    "The focus is to manage nitrogen after sugar beets," said Kasowski. "We are assuming that if we can get the nitrogen level in line right after harvest, we can keep it there for several years until sugar beets will be grown again."
    Centrol first implemented the project at the end of last year's crop. It involved 35 growers in 75-80 fields covering about 7,000 to 8,000 acres. Throughout the season, Centrol consultants walked the fields checking on conditions and making management suggestions as they normally would. As harvest time neared, however, the consultants began placing orders for satellite image acquisitions while the crops were at full foliage.
    The firm purchased SPOT FieldView VIGOR products from SPOT Image Corp. of Reston, Virginia, which for more than 12 years has marketed imagery collected by the international SPOT satellites. FieldView is a product line introduced last year and designed in conjunction with Agri ImaGIS Inc., Maddock, North Dakota, a distributor of SPOT Image products, specifically for use in precision farming applications.
    "The FieldView product includes the satellite images growers need to make detailed maps of their fields to define vegetative patterns of variability within a field," said Chuck Nichols, the Agriculture Program Manager for SPOT Image. "We also adjusted the frame size of the image to a single township so it's a very cost effective size for a grower or cooperative."
    The image is delivered on a GIS-ready CD containing georeferencing information so the grower or coop can register the scene with GPS points, vector maps, grid sampling data, yield monitor data or any other precision farming maps. Centrol loaded its images into MapInfo, a popular and easy-to-use viewing and mapping software developed by MapInfo Corp. of Troy, New York, for use on a standard personal computer.
    Understanding how Centrol used the image to manage nitrogen requires a quick recap of remote sensing fundamentals. The most important fact to keep in mind is that SPOT satellite scenes are not photographs taken by traditional film cameras; they are digital images acquired by sensors on board the orbiting satellites.
    These sensors measure the amount of energy reflected from small areas of ground, about 65x65 feet each, about 1/10th of an acre or the width of a spray boom, and then store these measurements as digital numbers. Each unit of ground surface scanned by the sensor is represented in the image by a pixel. Thousands of pixels comprise a single scene, and each pixel has coordinates relating to its location on the ground and a digital number corresponding to its reflectance value.
    Reflectance values are critical factors in the sugar beet project because the FieldView VIGOR product selected by Centrol is a scene composed of near- infrared reflectance values. Near-IR reflectance corresponds with the health, or vigor, of crops. Lush, leafy and healthy vegetation reflects strongly in the near-IR, while sick or stressed crops reflect less light.
    "When we viewed one of the SPOT images on screen, the areas where sugar beets had the lushest canopies appeared very bright, while less vigorous plants looked darker," said Kasowski.
    In digital remote sensing, when images are created from digital numbers, the higher reflectance values are represented as white or bright colors. Lower valued pixels are portrayed as darker colors or gray scales.

Corresponding Pixels to Nitrate Levels
In devising the theory behind the next step of the nitrogen management project, Centrol credits research by Dr. John Morgan of North Dakota State University. He determined the sugar beet tops, or foliage, left behind after harvesting contains specific amounts of nitrogen that is returned to the soil. The lushest or healthiest beet tops leave the most residual nitrogen.
    At Centrol field offices, technicians extracted individual beet fields from the satellite image. Each FieldView scene came packaged with a simple analysis software developed by Agri ImaGIS called AGI Viewer. This image was analyzed, and then exported to MapInfo to correlate nitrogen levels within a field.
    The AGI Viewer software allowed the consultants to display the pixel values not only as a black-and-white image, but also as a colored map showing variations in numerical values throughout the field. These variations corresponded to differences in foliage vigor and, therefore, residual nitrogen levels.
    "Ground truthing is critical in this phase," said Kasowski. "We had walked through the fields all season and we knew where the healthiest crops were."
    Digital numbers corresponding to the healthiest beet crops were assigned a dark green color. Progressively less vigorous areas were displayed in lighter shades of green until the lowest vigor zones were shown in yellow and red.
    Much of this process was automated which meant the technicians only had to assign colors to a few parts of the field, and the viewer software colorized the rest based on similar pixel values.
    "In just a few minutes, we had a colorized map of the field showing variations in sugar beet canopy vigor," said Kasowski.
    These colorized maps were then analyzed and manipulated in the MapInfo package. Technicians used MapInfo vectorization tools to draw lines around areas of similar vigor, which they called management zones. Then, based on years of working the fields and making references to Dr. Morgan's research, Centrol consultants assigned nitrogen credits to each zone based on crop vigor.
    "For instance, we figured the lushest areas would put about 70 to 80 pounds [per acre] of nitrogen back into the soil, the yellow-green areas 30 to 40 pounds, while the yellow zones would return zero to 10 pounds," said Kasowski. "We called these nitrogen credits, which meant we would put that much less nitrogen in those zones."
    Green management zones would receive much less nitrogen the year after sugar beets because the credit is subtracted from the recommended dosage, which can vary with whatever crop follows the beets. An 80-acre field was usually broken into about 10 management zones, some as small as a half an acre.

Applying the Nitrogen
The vector lines bordering each management zone along with corresponding nitrogen credit instructions were sent via e-mail to Centrol's Mapping Center in Mooreton, North Dakota. In the absence of e-mail, they could have been sent as simple TIFF files on floppy disks.
    The Mapping Center is equipped with data input software for a number of common variable spreader hardware systems. In Kasowski's project area, the variable rate contractor uses Ag-Chem Soilection spreaders. Ag-Chem Soilection GIS (SGIS) software was therefore utilized to import the management zone vectors and create the formulas to apply the proper nitrogen dosages to each.
    "The prescription maps are created right in the SGIS software," said Mike Freeberg, a Centrol crop consultant and Mapping Center manager. "The software converts this data into commands for the variable rate spreader."
    The Mapping Center outputs a color map of each field with application instructions along with a floppy disk containing the same information for the VRT computer. These items were supplied to Farmers Union Oil Co. in Climax, Minnesota, to carry out the application with a TerraGator variable rate spreader.
    Farmers Union in Climax has equipped its TerraGator with a SATLOC SwathStar guidance unit from SATLOC Inc. of Scottsdale, Arizona. This unit takes additional positional readings from satellites and combines them with GPS data to improve the accuracy of its nitrogen dosage placements to within 6 inches. Farmers Union Manager Bob Burner conducted the variable rate applications for Centrol, and although he knows the final results of the project will not be confirmed until after the next harvest, he is already impressed with the cost savings benefits.
    "What's surprising to me is that when grid sampling alone is used to manage nitrogen, more nitrogen is used, but with the satellite imaging, we put less nitrogen on those fields," said Burner.
    With the final results of the project still a few years down the road, Burner's point may be one of the most important in judging the near-term success of the technique: It can yield immediate savings for growers by reducing the amount of nitrogen applied to their fields.

Coordinating with Grid Samples
Centrol has been a major developer of grid sampling technology to take soil samples at numerous locations throughout individual fields. They use this technique to determine nitrogen variability in soils immediately before planting sugar beets.
    The firm plans to combine grid sampling and satellite image analysis techniques to further improve nitrogen management. Those fields that recorded very high nitrogen variability after harvest and received prescription doses of fertilizer will be grid sampled prior to the next sugar beet planting to determine if nitrogen variability problems have persisted.
    "Areas of low nitrogen will receive extra doses," said Kasowski. "Because we can't remove nitrogen, fields with too much will have to be managed differently, perhaps by planting a different variety of sugar beet that's more resilient to high nitrogen levels."
    Meanwhile, growers in the project area are confident they will see the project pay out in terms of higher sugar content in their next harvest because they already know the satellite imaging technique successfully identified problematic nitrogen variability in their fields following sugar beets.

About the author:
Kevin Corbley is a consultant and freelance writer specializing in remote sensing, GIS and GPS. He lives in Denver and may be reached at 303-722-0312 or by e-mail at [email protected]

Back