The Economics of
Remote Sensing for Management Zones
Jess Lowenberg-DeBoer
Introduction
A symposium on the role of remote sensing in the creation of management zones was held at the Agronomy Society of America (ASA) annual meeting in Denver last week. This symposium provided some insights into the state of remote sensing for agriculture. Essentially, it showed again that while there is some excellent agronomic and engineering research, commercial use of these techniques are growing, the economics are still poorly understood.
As background it should be noted that Purdue Precision Ag Services Survey in the spring of 2003 showed about 12% of US Ag retailers offer satellite remote sensing images to their clients. The Ag retailers indicated that about 25% of them expect to be offering satellite images by 2005. USDA figures showed that about 5% of corn acreage was managed with some type of remotely sensed image data (satellite or aerial photograph) in 2001.
The Purdue survey also showed that retailers are making progress in learning how to market images. In 2002, about 28% of retailers responding said that images were either profitable or covered variable and fixed costs. In 2003, that percentage was up to 55%. In both 2002 and 2003 about 29% of retailers said that images either were not profitable for them or that they did not know if they were profitable.
The ASA symposium included 16 presentations, with speakers from the Germany, Canada, France and Australia, as well as the US. There was one presentation on the economics of remote sensing (mine) and one presentation from a commercial provider of Ag software (David Waites, SST). The other presentations focused on agronomic interpretation and utilization of the data. Applications dealt with wheat, corn, sugar beets, potatoes, sugar cane and cotton. The list of presentations can be found in the meeting program at:
http://www.asa-cssa-sssa.org/anmeet/what.html
The symposium organizers plan to post the presentations in ".pdf" format on the ASA website.
Sugar Beet Experience
Dave Franzen, University of North Dakota, told the most commercially successful remote sensing story. He said that about 100,000 acres of sugar beets in the Red River Valley of the North were managed with the help of satellite images in 2002. The images were used to help determine management zones for nitrogen application on beets and carryover for subsequent crops. Beets are very sensitive to nitrogen. Over application can result in larger beets with lower sugar content and higher levels of impurities that must be removed in processing.
Franzen stated that the benefit of applying nitrogen by management is about $75/acre. The beet acreage receiving variable rate nitrogen (VRN) in this area is produced for a cooperator/processor, American Crystal Sugar. About 20% of the acreage that they process received VRN in 2002. The cooperative provides the images to growers for free and charges for processing to create management maps.
He noted that VRN beet acreage has been on a roller coaster. The practice was launched and became widespread in the late 1990s with grid sampling on relatively large grid sizes (e.g. 5 acres). Their experience showed that large grids were not adequately capturing the soil variability, resulting in reduced VRN management in 2000 and 2001. However, a growing number of producers have emerged recently with a strong desire to incorporate management zone approaches that involve greater detail and economic consideration.
Profitability
Because there is very little publicly available research on the economics of remote sensing for agriculture my presentation focused on how to estimate its benefits. One of the first steps is to quantify the objectives of zone management. Both remote sensing in agriculture and other zone delineation approaches have been plagued with ill-defined and unquantified economic goals. What little economic analysis of remote sensing has been done usually incorporates partial budgets to measure changes in annual cashflow. Only direct cash costs are accounted for, while analysis costs and management time are neglected. Further, benefits are often limited to the current cropping season, often excluding data and management value over time (i.e. long-term economic impact assessment)
The remote sensing and zone management literature is replete with references to improving environmental performance, but there is little agreement on how to measure that performance. Symposium participants debated whether environmental performance should be measured directly by pollutants in the groundwater and tile outflow, or indirectly in fertilizer efficiency and reduction in environmental loading.
Multiple goals are no problem for economic analysis. The classic approach is maximizing a weighted sum of the goals. For public research (e.g. to develop extension recommendations) the weights are often hard to specify. Often the next best alternative is to maximize one goal subject to achieving some minimum or threshold level of other goals. Specifying the goals in risk adjusted terms can include risk. For instance, the goal for zone delineation might be to maximize the net present value of benefits over time subject to reducing nitrate leaching to some specified level.
Once goals are defined, some economic issues that need to be examined include:
Remote sensing also raises some socio-economic issues. Are remote images particularly important for small or low resource farmers that do not have yield maps or intensive soil samples? If so does this imply a role for Land Grant university extension or the USDA in making and analyzing images for some groups?
Archival images may be particularly important for growers with newly purchased or rented land. The archival images may partially substitute for yield maps in zone development and other management choices.
The data needed for an economic evaluation of remote sensing for agriculture are becoming increasingly available. Crop and input prices are easily available. The agricultural image market is growing, and costs of images and analysis are available. Yield and quality changes due to zonation are harder to measure. Field trials are expensive and labor intensive, but provide credibility, while simulation and data mining are cheaper and less intensive to conduct.
With the accumulation of agronomic studies of zonation and other remote sensing uses the first step in economic analysis would probably be a meta-analysis which would do economic analysis by pooling yield, quality and input change estimates from the published literature. Simulation could be used to initially test zonation strategies that are not well studied in the existing literature. Field testing the most promising zonation methods with growers would be the ultimate test of benefits.
Conclusions
There are many good agronomic and engineering studies of remote sensing for agriculture, but the economics are not well studied. Available studies are mostly partial budgets for one season. Cost of images and analysis is often omitted, and management time is almost never included.
Economic issues include: the cost of the images and analysis, timeliness required, resolution needed, update cycle for zones, one-size-fits all zones or tailored to the input decision, the risk introduced by measuring proxy variables remotely, and the relative contribution of images to a zonation which also uses yield maps, soil tests, elevation, etc. Socio-economic issues include the potential for small and low resource farmers to use remote images, as well as the value of archival images for growers with newly purchased or rented land. The accumulation of agronomic studies of remote sensing raises the possibility of a meta-analysis that would pull together input, quality and yield changes from agronomic research, and estimate the potential profits and environmental benefits.
For more information:
Whipker, Linda, and Jay Akridge, “Precision Agricultural Services Dealership Survey Results,” Purdue University, Department of Ag Economics, Staff Paper, 3-10, June, 2003 (also available at www.purdue.edu/ssmc, click on publications).
Daberkow, S., J. Fernandez-Cornejo and M. Padgitt, “Precision Agriculture Technology Diffusion: Current Status and Future Prospects,” Proceedings of the 6th International Conference on Precision Agriculture, Minneapolis, MN, July, 2002.