Proceedings of the Workshop on Remote Sensing for Agriculture in the 21st Century

October 23-25th, 1996 
 

Session 1:  Who are the End Users and What Information Do They Really Want (And Are Willing to Pay For)?

Transcription Notes
 
| Ag 21 Agenda |
 
Moderated by John LeBoeuf
Session 1;  October 23, 1996
Founders Room, Alumni Center
Note:  Interjections are in [brackets].
JB=John LeBoeuf; P=Participant; R=Response
“...”= words muffled, cut off or interrupted by other participants.
Introduction
1. Who are the users?  Identify any major differences on who's going to use the technology.
2. What information and how much information?
3. Draw conclusions and make recommendations.

JB:  [How many of] you are actively involved in agricultural production?  We have a handful.  We need to know who is using the information, but we aren't limited to California agriculture.  We want to look at this on a large scale.  If there are differences in the southeastern U.S. or the Midwest with production of the bulk commodities (corn, wheat and soybeans), we could bring that into the discussion.  We want to look at some of the cropping systems like high value fruits and vegetables and also the permanent cropping systems of the orchards, trees, and vineyards. I'd like to identify the users and any differences between rain-fed and irrigated agriculture.  One of Susan's students and a Ph.D. candidate, Robert, will be working with us. We are taping this session, we'll be making some notes and trying to summarize the topic for presentation on the final day.  We'll give Dr. Susan Ustin a written summary and then they'll deliberate the possibility of a publication.

One of the key issues that we're working with at the National Research Council in Washington, DC, is this technology scale neutral to where any producer can use it for any type of cropping pattern?  There is a major concern:  will this technology allow the big farms to get bigger and push out the smaller growers? In California, we have over 250 different crops. We can't identify all of those today, but let's develop key areas to direct research and development, the use of cooperative extension, or the professionals that are in the industry.

Identifying Potential Users of Remote Sensing
JB:  Who is really going to use this besides the grower?  R:  You almost have to define growers.  You can have a major corporate grower that control a number of farmers.  JB:  When you talk about precision agriculture, we're talking about sub-field information.  We're looking about how the farmer can take that information and make a management decision for that field, whether it's changing inputs of chemicals or irrigation water in irrigated agriculture. We're not restricted only to precision agriculture applications. If this group feels there are end users out there besides a grower, then let's bring them forward and identify them.  Then we can go to the next level and say what information are they going after.  R:  Growers, consultants, marketing organizations. JB: You mean actual sales of commodities?  R:  Sales of the commodities.  We like to know what the growers are doing for others, but there other areas...  The legal area -- I would see lawyers paying a lot for this kind of information.  R: Let's do [growers] first and then we'll get to the lawyers.  And I'm not necessarily disagreeing that lawyers are one commodity but I'll put out the statement, "Lawyers aren't going to pay anybody for this right now."  JB:  Remote sensing is being used for litigation. But I don't want to discuss each one of these now.  I just want to get a quick list.

JB [repeats and rephrases participant's comments]: State government, such as Dept. of Water Resources, Dept. of Pesticide, Pesticide Registration (DPR)? What about EPA?  The federal Cal EPA.  We're going to see this as environmental issues come up.  Agricultural chemical dealers, distributors.   Cooperative extension.  Helping farmers.  On farm research? R:  Using the technology and bringing it to the people.

JB [repeats and rephrases participant's comments]: Irrigation districts?  Commodity groups was mentioned earlier.  Environmental organizations more than the state or federal level.  Consultants could be agriculture consultants, environmental consultants, environmental activist groups,  land developers.  I've heard talk about real estate people looking at this, insurance companies such as Rain and Hail for crop damage reports.

JB: I think there will be products that will go to certain end users and others won't find that useful.  You may find see some specificity in there.  There has been a lot of interest in the Midwest on corn, wheat and soybean on trying to see how remote sensing can help managers.  Then, you have to go to the other side and say well how can they take that information and make a difference if it is rain-fed and if they can't control the water which is a very important manageable factor in production.  If there are any other users, lets write them down.  The next step is what does each type of user need. R:  Commodities.  In litigation, you should have insurance.  Then, field rentals.  In the state of Washington, most of the fields are rented out.  There is not an existing market or user. [Muffled] R:  Has forestry been mentioned?  JB: I know there was some talk earlier that remote sensing being oversold to forestry.  I know there are some decision support systems that are being developed in forestry.  Who would be the actual end user?  R:  There are a lot of issues with forestry -- recreation management issues, water quality issues, wildlife habitat assessment, volume, overall planning, infrastructure management, water rights issues, mining, and inventory.  JB: Crop yield? Fire damage.  Sewage[?]Monitoring and recovery thereof.  Environmental damage (acid rain, ozone damage, etc.)  Rehabilitation (acid rain, ozone, Dutch elm disease).  Harvest scheduling ties in with forest inventory. R: Huge issue:  Stream management for fish management (regional scale).  Aquaculture relates to water quality.

JB: Any other issues on end users?  R:  Financial and lending institutions.  JB:  Good point.  Bankers, lending institutions.  I've also heard that in talk about trying to assess what quality of soil they have.  So, if they have a class one soil, we need to look at this in our rental payment.  Someone on the other side says, "Wait a minute, our remote sensing image indicates that crops have had a hard time growing her, so maybe we should not pay you what you are asking and lower the value of the payment."  There are two sides to that -- the one who's asking for the money or the rent, and the one who's going to do the renting.

JB:  Are the utility companies going to be using any of this data?  R:  Department of Water Resources and irrigation districts for the most part.  JB:  Energy?  Department of Transportation on the federal level.  State companies.

P:  One of the differentiations we have to make here, there are user groups that  absolutely must make a profit, there are user groups that don't care -- they have different sources of information at varying levels in the process.

JB:  Who's going to fill the need to determine what's wrong in the field?  Are the providers going to do that?  R:  I don't think there's a straight answer.  I think you'll have types of entities around the country that will fill that need on multi-levels.  Who has influence in Boise, Idaho may be one type of company, and that company won't have the same influence in the San Joaquin Valley.  Or it maybe a crop consultant group that may be strong in one area but not in others.  There's not one simple answer.  Geographically, those people will influence change dramatically.

Issues for Remote Sensing/Precision Agriculture
Frost Protection
P:  [Muffled response] We need to focus on where the money is ...?  Because some of these markets aren't going to matter.  JB:  They may feed off the satellites.  It may not be the primary target, but these groups may all feed off of remote sensing. R:  In California, I don't know any farmers who are using remote sensing for their irrigation scheduling.  We're putting a lot of emphasis on irrigation scheduling, and I don't see farmers wanting it.  But they do want to identify problems with their soils and problems with their irrigation systems.  I think the number one use for remote sensing in California would be for frost protection.  I've never heard anyone talking about using it for the frost protection.  JB: [and others]  We did do something with frost two years ago.  R:  The number one capital loss from an weather-related phenomenon in the United States is the frost.  I could see an application for remote sensing [to locate the] cold spots.  Is there something that is causing the cold spots?  Is there something we can do to change the topography that will eliminate it?  JB:  There are insurance companies looking at that and the topography and using that in a GIS where they can layer topography and then pull up weather information going back 20 and 30 years and looking at potential sites for frost.

P:  Is that kind of a one-time situation we're trying to define?  Pockets that are frost prone?  JB:  They'd like to know before and after also a frost  P:  Somewhere on the line, we need to be thinking about plantings.  If the frost protection is a one-time, now you know where those frost pockets are located -- you don't have to go back and do it everyday.  Other things are very time dependent as to whether it's useful information or not.  You have to have daily, weekly, etc.  JB:  On this topic of frost, I know there are people trying to see if they can get  a before and after image to assess crop damage.  With citrus, the producers need to know if we've had a frost, where have we had it, and then let's go out and send a harvest crew there to pick produce off within a day or two.  Then they do not have to worry about having the quality of the fruit go down to where they can put it into lower priced juice, instead of the fresh market side which receives higher premiums.

P:  These guys thought they had soils problems and irrigation problems.  As it turned out there was cold air draining, so a lot of these trees were smaller.  Then they put their smudge pots in those areas.  In frost irrigation, you have real-time stuff ... Did that irrigation water do any good, running 2 acre-feet of water that's freezing cold.  Put some fertilizer on this citrus orchard.  The other things we did we had guys sitting there that would say its 28 degrees Fahrenheit, but the temperature of the plants and the trees was still 33 degrees.  So you initiate frost protection techniques right now so you can delay that and you're looking at thousands of dollars per acre.  On the other hand, we have the insurance.  We flew and knew where all the trees were healthy, wait for a damage person to come and assess it and then ... they tell you to take everything out to above 4' above the ground.  Then you have re-growth.  Now that orchard will grow for next year, but what's it going to do this year?

JB:  This is where on the permanent crops, I see a lot of value. It brings the analysis in on remote sensing.  Before we've had detection and recording, and it's been very weak on the analysis.  Now we want to be able to see how we can bring this in and beef up on the analysis of the information and get that into the GIS and start looking at layers (fertility of the soils, topography, etc.).  P:  Are we limiting our discussion of remote sensing just to imagery?  Because I work with a company that is desiring to set up a weather station every 5 or 10 miles throughout a county and it gives real-time weather feedback that is related to ... remote sensing package. R: I think we want to focus on all of these technologies together which are going to change the way agriculture operates. R: With that in mind, then there could actually be an interpretation of all these bits of information that really is what the end user is using.  What I keep hearing is, if there's a need that's economical, then there's a market.  If there isn't a need because not enough research has been done, it doesn't matter because of other limitations without a market.  That's the common thread to all of these concepts.

Irrigation Scheduling
P:  When you have irrigation dependent crops, we water everything.  If it rains, we water anyway.  In the Midwest and Louisiana, people have supplemental irrigation.  Now there's enough research out there that you know if the cotton crop has gone without water for 34 days, I'm going to lose 40% of my yield.  If I water after 18 days, I'm only going to lose 10% of my crop.  That's more of a supplemental irrigation.  There are other areas that are just dry areas.  There are areas where there is no irrigation.  You take what you get.

P:  Let me give you an example [of a project] in California set up for irrigation scheduling.  I talked to one cotton grower in the San Joaquin Valley.  His daughter went to Fresno State and she programmed all the irrigation systems to water at certain times.  He had hired the largest irrigation consultant in California at that time.  On the side, I asked him if his irrigation consultant was still working with him?  He said no.  He found out that it was the same every year so why should he pay the irrigation consultant if he irrigate the same schedule every year? I calculated out what he was putting on relative. The first irrigation, his efficiency was terrible because he over-irrigated and the last irrigation his efficiency was close to 100% because he under-irrigated.  It's a complicated mess and he didn't have a lot of control over it because he had to irrigate when he had the water.  We don't have rain in California so he didn't need to worry about that.  If you were going to irrigation scheduling, the only place I could see a real good use for it would be the winter cereals of wheat and barley in California.

P:  I was just going to add onto that.  The Tulare Lake Basin -- up there we ran into irrigation situations where we thought if we told them to irrigate, then they'll irrigate.  Well, that's just not the case.  If you have 5 fields, and you're irrigating just one field a day and three fields are stressed, there's nothing he can do about it.  He's got to irrigate so many inches on this field and then switch them.  So, there is a  complexity in the market and it doesn't provide this information, one of the caveats of the customer being able to get value from that is that he has to be able to do something about it.

JB:  That may be if the stress is under-irrigation and you're trying to schedule this but there is remote sensing being used for detection of stress by over-irrigation.  Then you have some time to go back and do some ground-truthing and say we have too much water here so lets go back and adjust our irrigation schedules.  I've seen this in the San Joaquin Valley so I know there are people looking at this.

P:   Well, ... I think the key element is , if you could do something about it, if you had a drip system, and you can turn it on and off, that's information ...  If you basically depend on the water district, they're the ones who control the flow of water within the district.  Now one thing I think you'll see in the Midwest and you'll see it in California...  will be a type of information about the impact of weather.  I see a lot of times one has to ask ... How much does it affect the yield?  5%, 10%?  How much does it reduce the risk of getting  the crop into harvest?  Using weather information can save the crops and make a difference.  If you grow wheat in Kansas and you make ... 5%  What does a user expect to pay for information versus how much money he will make ?  If you make $10 or $20...that's a difference worth the investment for information.

Differences between Rain-fed and Irrigated Agriculture
JB:  In California, irrigated agriculture has a big plus here with the irrigation districts because we manage our water and our inputs.  R:  I think there are a lot of different dimensions [between rain-fed and irrigated agriculture].   I don't think this group should try to find a solution for all agriculture.  The distribution information will have to be on a local basis, because things are so different.  For example, how long the season is, and the size [of farms].  There isn't a simple end user out there.  JB:  What would be different from a rain-fed user standpoint?  R: Here's the basic question from the end user, "You gave me this image.  What would I do differently?"  If it's just a pretty picture, that's entertainment.  Is it something that guides me to take an action based on what you've given me?  Then I have to figure out if I make money doing it. JB:  Do you think sensors will be different or not?  R: Let me make a separate point.  Each area and each type of farm in a region will have different demands. I also think that remote sensing, a piece of it would be one component of information they need, but I don't think mapping that image is enough.

P:  Another approach to what he's saying.  There's a common need across the country for irrigation management.  We do a lot of irrigation management in West, but the needs in the Midwest are a little different than they are in California.  The basic fundamental problem with irrigation [between both groups] is conceptually the same.  There are needs for using remote sensing data for logistics on moving people and machinery around in the most efficient manner.  How they do it in California is different from how they do it in the Midwest, but the fundamental process of using the data as a human resource management tool is common.  We [in California] are going to use it for disease prevention. JB:  [End User] Decision Support Systems.  R: And consultants may use it differently for diseases than a farmer than it is in a larger agribusiness, etc.  All of them are going to be using some aspect of the same fundamental piece of information.  R:  If you have the irrigation, usually you have higher value crops [to offset the cost of irrigation] to get the most amount of money.  R: You have more money to make different kinds of decisions. R:  We've got a lot of high value crops (sugar beets, potatoes, corn, etc.) in the Midwest.  JB:  Yeah, we also do wheat and barley here in California.  R:  One difference is that the southeast has more diversity.  Fields are smaller.  You may want to have the ability to identify one type of plant/crop from another.

P:  I think we have to decide what the end product is and that will define the end users.  We have a whole continuum of end users in the process.  If you're the remote sensing vendor, you're just selling the picture, your end user may be a research company who takes that picture and makes it meaningful in interpretation.  They may not be the people who, in turn, work with the growers one on one.  There's a series of products that are delivered through the process of remote sensing. Then each have their own market of end users depending on what the product is -- whether it's a broad picture, an interpretation, the implementation of that information.

Stress Detection
JB:  There are a lot of things that remote sensing may be able to pick up on plant health. I really think that remote sensing's strength is stress detection --whether it's water stress (over-irrigation, under-irrigation), fertility, plant disease, pathogens, nematodes. I think a lot of the value of stress detection is if you can detect the problem and use it as a way to map the problem, then you can use that the following year or the following profit season, maybe a manager can change his variety or he can change his fertility practice or irrigation. Are there any topics here on plant health or anything that you feel as a group that is really going to drive remote sensing into the adopter’s hands?  R: Weed infestation.  JB: We've heard about research in the Midwest where they're looking at using remote sensing and mapping different areas and different problem weeds in the small grains.

P:  Frequently, before you have an insect infestation, you have symptoms.  If we have some way to provide previsual detection of an insect or disease outbreak, we've been looking for that for 20 years.  For instance, bark beetle.  Some folks feel that there are some promising areas and thermal sensor stuff to help us do that.

Maturity/Harvest Times
JB:  Harvesting time -- we'll put that in with maturity. In our melon production here in the San Joaquin Valley, we're looking at that -- if you could follow the digital information that goes with the remote sensing images.  When does a crop peak at the maximum vegetative growth and start to come down and hit senescence? When fruiting structures start to form there is a change in allocation of nutrients.   Where you have the energy that's stored in the leaves that then goes into the fruit, there is a physiological change from a sink to a source.  We have talked about crop phenology models, we're looking at that.  A lot of digital information is really trying to key us in on that approaching maturity.  In some of the crops you may be able to use that information to help your marketing of perishable commodities and that you can tie in with your marketing statistics.  A lot of the food chains will contact the producer and tell him, we really would like to buy your produce and put it into an ad three weeks from now we want to buy X amount of this commodity.  If the grower can get an idea on when a crop is maturing it will help marketing and will impact your processors, packers, shippers and food chain distribution. In contrast, the commodities that go into storage, that won't play as much importance.

Identification of plant species and vegetation cover
P:  Even the species classification is part of it.  JB:  That's really going to be dependent on the resolution.  I don't have enough background in forestry to even tackle that.  R:  I have to agree that our most overriding requirement of remote sensing would be the vegetation cover.  And we've taken a multi-resolution approach to that ... for site specific areas, use high resolution.  JB:  Any difference there on the scale?  You're saying on a regional scale or on a smaller scale?  R:  Yes, for regional scales, we use coarse resolution satellite systems that cover broad areas.  JB:  So is the need already being provided?  R:  For site specific, high resolution area, special project areas, yes.  We need high resolution data, which is not being provided for large areas.  We'd like to have one meter data.  R: But how often? R: We don't need it that frequently.

Application of biological control to problem "spots"
JB:  This is where biological control will really be enhanced by pinpointing areas.  If there's a spider mite infestation in a field on a small scale and a grower has a field scout available, they can order predacious mites and send the scout out with the DGPS unit.  You can take information on a smaller scale and incorporate beneficial parasites or predators into that situation and enhance the management, rather than waiting for it to flare up so big that you need to call the plane in for a chemical application.  Then you go back into the chemical application side, when you have a threshold for chemical control rather than the biological control.  Good resolution will help in small scale applications for permanent crops.  With insecticide applications, the farmers say, "Don't tell me about it -- let's see if the beneficial insects take care of it.  When it threatens 25% of my crop, I'll push the button to call the pilot in."  There are some differences in pest management that we're going to see a change in the information needs.  Biological control programs of augmentation want to know site specific areas as soon as possible.  When remote sensing is used as a crop monitoring service, images will allow the user to see changes from week to week.  If you go in and just take an image intermittently, you're not going to catch these small problems.  R:  If you get an image only every 16 days, you may miss the entire infestation.  That's why they want them on a regular basis.  We've gone twice a week -- three times a week is too much.  If they try and cut cost and only look at a map once every other week, I show it to them and ask them if they can discern problems in their field.  Invariably they always realize that they need the maps more often.  They need to plan their schedule around the Monday/Thursday maps.  First you can get water, then they care about bugs, fertilizer, weeds.  That's all they care about.  Now ASCS is going to give them their acreage, and they won't have to worry about that.  Now the insurance company may want to use that map over again, but the grower doesn't really care unless he has hail damage.  Then he's going to take your map, look at the hail damage, quantify it.  When his hail adjuster comes out there, he's got this map that covers 2 square miles.  That's not going to be the driving force to pay for remote sensing, because it's so infrequent.  You've got to the market the product to fit the needs of the day to day grower's operations or the consultants concerns.

Use of Raw Data for Miscellaneous Needs
JB:  Do you see a lot of the consultants asking for the raw data themselves? P:  We have [had consultants ask for raw data]. We're dealing with thermal predominantly.  Thermal is a difficult thing to understand.  An interpreted version might be something which would work better.  We provided a thermal image  but we did not interpret it.  We wanted to give it to our clients for a year.  We're working on our Crop Water Stress Index and other applications, and we wanted to let them see it for a year.  We would point out correlations.

P:  That's the first time I heard that 25% threshold.  There's an inertia, because it has to be monetarily worthwhile for them to do that.  On a low value crop, we've shown pictures, we recommend this.  But what will it cost?  They can't do it, because the margin isn't there, so they have to leave it.  On a high value crop, they can do something about it. So there's a balance to the urgency that they have.  When you tell them they have a problem, they have to have the apparent capability to address the problem and that relates to technology.  They didn't have the GPS crop duster.  Once we start getting the data sources at finer resolution, people will start coming up with the technology -- the GPS driven crop duster.  The Japanese have this helicopter that's precision GPS.  You can just spray one 12' x 12' section.

JB: What I'm trying to get at is if we see a need for the digital information.  What I perceive is that there's a real need for the analysis of the information -- just giving a color enhanced map to someone is one thing, but if you actually get the raw data inside it and then start doing the analysis -- by analysis I mean, looking at State Extension personnel starting to analyze the information and  bring that into crop phenology and prediction models.  If we can identify growth stages and know areas where we can control it (like in irrigated agriculture), where we can push or hold back the inputs -- if we have enough information to determine whether the crop is tracking how we think it will.  R: We have many applications.  Now in computers, almost all of the farmers have adopted it to do really basic stuff for accounting.  A small number of people use it for irrigation management, etc.  I think you need to identify the components ...it is difficult to focus on digital information ... small sample size and all the concerns... When you speak about something, what is the critical point -- what is it saying to people?  Slowly, once they get used it, the technology will be adapted.

P:  A lot of that is also time sensitive... one or two growing seasons would be fine, but now what do  I do with this data.  What was my yield?  What happened?  Am I going in the right direction?  They won't use the digital radar right away, but later on there will be a request for longer term trends.  JB:  I see that having value on the farm on a field level, but I don't know if there would be value in the digital information on a regional or global scale, because it's such an immense dataset.  Who's going to analyze that?  How are you even going to store that and access it?  It's something that if the research community could access that (there would be a need to develop infrastructure first) and could handle the data and see where are there applications, maybe that would have a lot of value there.

JB:  So the need is more for the people who are looking at a larger scale?  Is that what the group thinks?  That the farmer or the field level analysis won't be as strong as some on a regional scale?  Even in the commodities, will they be able to access information crop specific and then analyze that across the season.  One of the things brought up about the use of information:  A couple years ago when you had a cotton crop failure in Pakistan, India and China, it drove the price of cotton up to the highest price since the Civil War and the producers or the people in coops who knew that something was going on were able to handle the situation and keep their cotton in storage for several months -- they were literally making millions in a short period of time compared to the growers who put their cotton into the market right after harvest.  A lot of time there's a 40 cent difference of each pound of cotton.  So on a large scale, there's a lot of value there.  But what I'm gathering from this group is that digital information will be very limited in its use on the farm scale and may be utilized by the consultants or advisers, but not by the actual grower or producer.

P:  I don't understand what you mean by digital information.  JB: Let me break this down into an example of a  Resource21 image that they're marketing in California.  You would have a breakdown of 20 categories of biomass vegetation index.  The statistical information brought to the grower included standard deviation, range, minimum and maximum number.  As a consultant, when you start plotting that across time, even the maximum numbers from week to week to week, you had crop growth curves that were already being identified in crops.  Once you hit a peak and your crop hits a maximum, then it starts to come down.  That's where we tie in with the key to find information to help identify when is this crop going to reach maturity?  When do you have to stop your inputs?  As a producer that was using this on a field level, I found an incredible amount of information in the raw data.  Then we've even taken the information to the GeoInformation Technology Center in Fresno State and had them break the raw data into percentiles and plot the percentiles across time.  It all followed the same pattern.  Your canopy was reaching a maximum, then it hit the top, then it started to go into senescence.  This is where we're trying to identify the period of time from reaching the maximum number in the vegetative index until crop maturity.  For example, two and a half weeks from now, it's going to be ready to harvest.  The key area I see in irrigated agriculture is that the last irrigation is usually the most critical and I see more damage to fields done by over-irrigation late in the season because the grower gets nervous.  You impact plant diseases if you put too much water on it, your quality of your fruit goes down.  We see that in produce where the sugar development decreases in melons if you over-irrigate.  I think there's value in this and it ties back to the irrigated agriculture, because you control the input.  I see value in the raw data as information within the information.

Georegistered data to identify field boundaries and roads.
P: The number one request I'm getting for remote sensing today is for high quality, geo-registered data that I can use to draw my field boundaries.  It's really a one-time use, but a lot of people are implementing GIS systems in the coops.  So they now want a high resolution map so they can go in and draw all the field perimeters in the office, rather than going out and driving the field.  JB:  So that's a single use.  R:  Some of the guys who've wanted our maps, only because it's a good road map, so their guys didn't get lost.

Soil Sampling based on Visual Factors in the Landscape
P:  As they get familiar with the data and its uses, the second thing I hear is asking for soil sampling, volume.  Now they don't want a grid sample, they start sampling by landform or some sort of visually definable differences in the field.  To define slope, drainage areas, hills. Grid sampling has evolved into sampling by visually discernible differences within the field.  As we look at the evolution of remote sensing, 1) base map perimeters to link to a customer base, location.  2) Soils information.  JB:  Do you see the need for soils mapping more than once a year?  R:  No.  Most of it's once in a lifetime.  But these are the processes that I see going through.  Once they're comfortable, they'll move on to 3) looking at the crop during the summer season, for example.  Then it grows into a lot of the things we've talked about [that require more data requirements on a consistent basis].  R: You hit it on the head when you said road maps.

Salt Water and Drainage Problems
P:  Down in the Imperial Valley, we have salt water and [drainage problems].  R:  You've got the other problem where you have too much water, so you start identifying areas where you compile and manage the water.  But those are things that I see coming.

JB:  On the soil color maps that Resource 21 was looking at, is that something you're seeing?  Are there any differences between what people in the Midwest are using it for in rain-fed agriculture?  Is there anything different or any different value that you see in soil color maps?
 R:  I think there's just a little different twist on it...  R:  ... periodic evaluation of Landsat data gave us a pretty good feeling for an area that was growing well with nothing wrong and other areas that were going to be a problem that didn't look right.  That's where your concentration goes.  It really helps you out to maximize the utility of those fields where there are problems.

Introducing the Technology and Discussion of Market Structure
P:  We had some examples of this kind of a problem in the past.  The introduction of computers back in the 1950s.  The introduction of new technology has a very expensive tag.  In trying to identify who the customers would be, you have to ask who can make major investments.  We looked at the companies in the Fortune 500.  There was a management problem that came up with respect to how they would manage this major investment.  One approach is you charge everybody exactly what it's going to cost as best that you can determine.   The other approach is  an overhead approach to find out who can use it and how they can use it.  I've seen this happen in at least two areas -- the initial computer entry and the second was PCs.  When PCs came out, it was the same kind of thing.  I knew people who decided to buy 10 computers and put them in the company.  [They didn't know how they could be used, but were intent on learning how to use them].  So it seems to me that you can't avoid the question of who are the people that can make the investments to introduce this new technology to the ag system?  R:  Those are the innovators.  The ones who grab new technology and they may not even know all the areas where it can be used.  R:  In the early computer days, it was the board of directors and the president of the company who made the decisions.  Those are the people that you had to talk to.  R: The people with vision. R: And money!

P:  We don't need to list every end user who will buy remote sensing, because they're figuring it out for themselves.  The main point that I think I'm hearing is that we've got to narrow who we think the end user is going to be.  We can list [a lot of people], and you're going to have to hit regions where already the PC has moved into [the farmer's] standard.  Not in every part of the country.  In an area where we are, about 36% of farmers actively use PCs in their business.  That's a ripe area to start walking in with a new technology.  Are they at least [computer] literate enough to even have the discussion with you?  Once these yield monitors appear, and that's happening fast, that technology is being adopted.  The gap that remote sensing could fill is "Here's some information during the growing season, and based on that you can do something to optimize your yield or maybe you can't do anything and save your money --whatever, but you've got information during the season that you could base a decision on" -- those people are willing to pay money.

P: If you, for example, have a $10,000 piece of equipment that can gather field data... if you sell the information  instead of the sophisticated equipment...If you start asking users on a farm use basis... operators won't mind $2,000.  A lot of it really depends what type of management strategy you want and what type of secrecy is important.  It's clear to me  there must be a return on capital investments... to start saving capital especially, but beyond the yield monitor, then you really have a tough question.  You basically try to take one technology... try to go in and show the value that will lead towards another.

JB:  There's a very wide range of means of getting CIMIS information, and I think some of the usage has been a little limited, because they're looking at who's on line with it and who's tapping into it, but it's in the media, in the newspaper and on the radio.  There's a lot of ways to access it.  R:  Actually, if you'll look at this from a different perspective.  Who's really going to be able to control and distribute this information?  In most cases, the information is not going to go straight to the farmer, it's going to an information multiplier who works with it every day and can afford to do it.  Those are the real customers.  Those are the ones that have the money, the infrastructure and the process.   They develop the infrastructure, mold this process, interpret it and add value to it.  Those really, in many cases, are the customers.  The farmer is paying for it, but they're purchasing a service from somebody [else] and that's  very critical.  You're looking at large consulting operations.  You're looking at a very limited number of people who are going to be introducing and managing this technology in a fairly large scale.  There's going to be lots of individual exceptions to this.  Let's take the whole ag input group -- all the farm suppliers, etc. -- 80% of the U.S. market is going to be controlled by 15-17 different companies -- they are going to be the ones to manage the input of this to all of their customers.  Now let's take crop consultants  -- if you take the whole genre of crop consultants, I bet that there aren't 40 or 50 in the country that could implement something at this scale.  Most of them have barely gotten into laptop computers.  When you look at the processing companies, you're dealing with a small number of entities that have ...

JB:  Is this the comment that you were making earlier about the agronomic side lagging on being able to grab this information and use it?  R:  We don't have any trouble with technology.  Acquire data, process data ... The problem that you're going to have ... on a daily basis ... and I think the key in all of this, whether airborne or satellite systems, you're still going to have to realize the grower is the one who's going to have to pay and the only one who's going to use the data on a consistent basis.  The commodities people need an overview of it.  The processors need to know how much is in the ground, but he doesn't need to know weekly information which all of these systems are built to provide. The driving force is his prediction to that information.  Yield monitor data and he's got a GIS, but he doesn't have anything in between.  He's got to fill that gap.  Remote sensing provides him with that information on a timely basis.  P: The person who pays is not the person who interprets, that's what you're saying.  It's the agronomists and the coops, ag supply people who are actually making the information useful and interpretable for the grower, the end user.  So it is the end user [grower]who  is the person who pays and is the person who's using that information  made from  interpretations.  JB:  There are people that are consultants who are purchasing remote sensing images and then taking that to the client, as part of their service.

P: We saw that there was an existing business structure out there.  It was also a location that had agronomic expertise.  This is connected to do our actual customer service.  The money comes out of the grower's pocket for something that's going to directly benefit him.  So I think that's the recognized product.  Now the coop's don't do this for free, so they may make a little money out of this too.  So there's a little bit of change that goes down along the process.  You might want to give the coop raw data, the coop may have the expertise to develop that into value-added products for specific customers, so they make money out of the value-added.  If you want to sell value-added products directly to the customer, it's going to have to go through the coop.

P:  With us, we have a business that we sell to the growers.  We make maps and quantify those maps.  The grower doesn't care about getting a color map.  In essence, you could say that we are the consultants.  He gets that on a regular basis, and you want to make him dependent on it.  Then I've got guys that want acreage.  But you can't run a business based on acreage.  We can do that stuff, but it's one time for maybe 10 cents/acre.  The trees, once you know where they are, who cares?  What I'm getting at is that once you get that information, how often are you going to need it?  What is that satellite going to do the rest of the time?  You may not need it again unless you get a forest fire or clear cutting or something like that.

P: [barely audible] We've got a number of different points...  You may sell that once ... lower, but you may have an insurance company that wants it five times ... or a financial institution that wants it three times ...  Remember also on the chain... instead of it being 10 cents which you've got to pay the government for ... 20-50.  It's got to be consistent.  If it's not that consistent, it's worthless.  You can't get an airborne program to run ... you miss your insurance payment, etc. and ... then they say, "That's it, this has no value to it ..."  What are you going to do?  Real high resolution day in and day out kind of stuff.  You're looking at high value crops.  Obviously you're not going to do day in and day out flying on non-irrigated wheat...  So they want that first year agenda ...  Maybe it's not a U.S. issue, maybe it's some coop from Mexico that has to get over here or there,  but they submitted a claim for ten million dollars yesterday.  What do we do? ... They sent him an $8 million bill.  R: Another  example on that is a number that you know as well.  Something about in commerce we look for specific things, in the code of regulations.  ... not money to monitor what really is going on in the field.  P:  I've got to look at the people who use technology , ... how much money will they spend ... litigation for they have to pay for CIMIS information.  Exactly, if you had a $10 million dispute, ... certain condition, location, time, you'd be ready to pay.   If you have a big customer ... it seems that they utilize technology - these guys wouldn't be the one that needs help in using information.  What you really have to identify is what you have  to market... that really puts some of the foundation and support for development of  information.  You really have to try to look at your options, and this issue of time.  ... information every day.  If you look at the end of the week.  If you look at some of the information needs for Western agriculture ... in California, you may need two or three times  a week... identify some of the disease or insect infestations or other things.  I think, for example, that people would be very lucky to gather  information ...a  certain amount may be beneficial ...of some sort of value... I see you would be in a much better situation to make a management decision... If you can provide information that would be variable at the frequency of different types based on the system.  In other cases, it may be three or four times in the season.  Maybe for other individuals, pay extra if you can get it on demand at certain times.  A lot of the demand depends on what kind of information you're using.  I think the key issue is what type of information can help growers ... You would be able to get at the information that would be relevant.

Products
JB:  Some of the concern in the agricultural community is they're going to get too much information.  I've heard so many growers already talk about information overload and that they already have so many decisions they have to make.  I think if we could determine, what kind of information are they trying to look for?  What do they actually need?  If they can't use that image to make the decision or help in a  decision support system (DSS), then why are we doing this?  R: When you're talking about giving them a map, we need to differentiate between an image, an enhanced image, and a classified image. Do we provide a raw image with raw data or do you provide a classified image?  R:  We provide a digital map that has been processed, cut down to soils, root zone and plants.  Then that .. is processed ... a digital color map that matches the report, in this case water stress.  We're looking at fields and we look at each field separately.  First we measure the water stress value for each field.  Then that's basic research ... in this case irrigation scheduling.  You don't want to stress  the crop... but you do want to stress cotton a certain amount.  So then the grower ... we would give them the color map with a quantitative report.  The report quantifies the whole field, the map shows spatial variability due to bad soils, for example.  We have ten colors.  Each indicates a certain amount of water stress.  Once every three years, someone will call up and say, "Do you have a map of the storm damage out there?"  That's value added, but it doesn't drive our business day to day.  JB:  Are you saying you're giving [the growers] that information on irrigation, when to irrigate?  R:  ...We fly, get an image of the farm, the fields are numbered, they know where they are and we have a correlating report that quantifies that.  We tell the grower 80% of ..., 20% is what you really want if the threshold for irrigation of cotton is 0.3.  He want's to know it's 0.2 today, when do I water?  JB:  You're leading them into a decision?  R:  That's right.  In the past this last year, we took the decision making process off the cut and cost.  Now I have guys coming back that want to pay more to get the decision making process on.  Then they give these maps to their irrigators, and the irrigators never look at the report.  They took those maps out and that's how they do their irrigation.  Now they can tell where the hot spots and where there is standing water.  So they mitigate the hot spots first.  JB:  That's getting site specific.  R:  And that's what these guys are doing... That's the driving force ... JB:  This is where the vegetation change maps we've seen in the industry can kind of enhance what the people are thinking as producers.  The vegetation change map shows one week to the next .

R:  That's right.  In Australia, they have heavy clays like you have here.  With them, they have problems of water logging. Water logging can cause more damage because of water stress than under-watering.  There he faces a problem, the U.S. tells him you want to water at 0.3 -- that's where the highest yield threshold is when you're input water comes.  It may not be your best cost, but that's where your highest yield is.  Down there it's probably 0.45.  They want to go much further, because of water reasons.  If you water too early, it might raise the water stress up to 0.6.  What I've seen is that the grower don't want that -- we've tried that a bunch of times.  They don't want anything for free.  If you give it to them for free, they won't use it.  [The growers are] looking for a value-added product where you give them a recommendation.  You have this map, now they need a recommendation.  General mapping for insurance purposes, they don't necessarily want [a recommendation].

P:  Some people want the added value data, some people just want the broad data that says tell them where to go look.  It doesn't matter what the problem or stress is, but tell them where to go so they can scout that area.  JB:  This is for all levels of the people on the ranch, whether it's the consultant, the field scout, advisor, or the ranch manager.  That's what I see as a strong point for remote sensing -- being able to identify these areas so that people can go out and spend less time in the field.  R:  And whether that's done with tractor base or equipment base or fly-over or satellite sense, the farmer wants it in the most cost-effective way.  R:  That's our view of the market too.  One guy just wanted the piece of paper [not with any value added] for good reason.  They didn't want Mike McDonald, who might never have walked through a cotton field before, telling them where the bad spots were.  So there's a market for both types of information.

How much are people willing to pay and who's going to implement remote sensing?
P:  Coming back to your original questions in this discussion, what's driving the use of remote sensing.  What's hindering it is a lack of people or organizations that can take the information and generate recommendations/actions quickly and efficiently that are financially beneficial to the grower.  How much will they pay for it?  They'll pay a little bit less... R: $10?  The providers want $20.  R:  So if the information of one value is beans versus corn vs. grapes, there may be different values and they may be willing to pay different amounts for that information.  Regarding who's going to do it, it's whoever can assimilate, interpret the information most quickly and efficiently.  R: The independent group that could do that is cooperative extension.  The growers would like to get the data from whoever -  NASA ... because it would be free to get the data and to learn the use of it, and then to translate it to the people.  We start with the specialists at the university level.  Once they know enough about remote sensing, they'll teach their county and extension agency. R: If I ran my company in your state, I'd sue you for competing.  R:  You'd hurt the companies out there by giving information you got free to growers for free.

JB:  If we look at the cost, and the grower wants to spend $15/acre for technology, most pesticide applications are $11-20/acre.  But if you can treat a portion of the field instead of the entire field, then you're getting money back.  The cheapest chemical is dusting sulfur.  We've been spending  $5-7/acre/year for technology and that's not even equal to one dusting sulfur application.  If it gets above $15-20, then you're going to the range of pesticide application where farmers don't want to spend that kind of money unless they have to.  There's a cost factor.  This technology will need to be less than a  pesticide application.

P:  I'm working in a company that works with private industry and cooperative extension on implementing a new technology, which is land based.  Cooperative extension has been more effective in starting research, however they're not profit-motivated... So there needs to private industry to generate the profit for them, and there needs to be some public or private research organizations... R:  I was an extension specialist for seven years in Arizona.  We went out with hand-held thermal devices and showed it to people.  The next year we came back with plots that showed yield.  By the time we put thermal in the air, they were all sold on thermal.  The opportunity with extension is that you can demonstrate [technology] to the growers. R: It's a teaching, education and consulting role.  Extension is not going to make the value added products.  But they are going to help their people in choosing products and making recommendations.

P:  There's a lack of information.  I don't know what's available. R: This is where space grant can help... agribusiness together with farmers and the cooperative extension.  R:  90% of the space grant people are resident in engineering and aerospace.  So they'd have to reallocate their time a bit.  R: Bring the farmer in, explain to them how to interpret the data and explain the limitations and give them some confidence.  They play a different role than a private company, because they aren't selling the product... unbiased R: The role of space grant ... mostly education, scholarships etc.  There's been a shift in the last two years with more emphasis on the extension side.  JB:  I think the key word I heard was "unbiased."  The end user would like to know they could have confidence in who was giving them the information.  It's an outside source that's not trying to sell them something. R:  I think it's a process.   First the user wants to use the technology.  Then they go to get advise on which [technology/product] is best.  Once that occurs, the training and implementation process needs to be done to serve them with the speed and efficiency that  is provided by  the private sector... R: That's exactly what I was going to say.  If someone came to us asking about remote sensing, we don't have the time to explain everything.  But if there was a home page that listed all the companies and explained all the options and services...  R: One of the missing gaps is not only who do I go to, but is it for real?  What can we really do operationally versus what we're hoping to do?  We need to get that pinned down, so we give these potential users realistic answer.

JB:  One of the key roles I see in the state extension is it doesn't matter what size farm they have.  So that eliminates the issue of scale.  It makes this technology available to everyone, not just the large farms or high value cash crops.  They can draw on state extension as a resource, instead of having a company saying they won't service the grower because they don't have enough acres in production.  R: ... Remote sensing funding [should] go to state extensions to provide them with the utilities to introduce the public to some of these [products] in a general way, not one over another.

Threshold Necessary to Implement Mitigation
JB:  One of the key thoughts that I've had with talking to potential users is  what scale of stress in their field would they respond to.  In general terms, a lot of the input that I've gotten from managers is 25%.  That they're not really going to change anything unless they see that there's a need that's threatening 25% of their field.  It doesn't matter whether the field is 100 acres or 40 acres.  One of the needs I'm hearing about is where is the threshold where information will help make a decision, rather than just saying here's a problem.  Once a farmer is on a field for a few years, they figure out where their hot spots are or where their drainage problems are.  Remote sensing may enhance it, but they already know it's there.  If they can't respond to it in a management decision, then it's not going to have added value.  R:  If you have predictive value, you can tell them a week before a hot spot will get hot.  The nice thing about remote sensing is that you can see these things [whereas in the field, there aren't any visual signs].  Once cotton wilts, it's way too late.  With spider mites, if he's only got it in two acres to begin with, he's going to have to do something otherwise it'll spread and he'll have to spray the whole field.  You can give him coordinates to spray.  Now you're looking to mitigate things and [save the farmer money].

Justification of Remote Sensing
Yield Monitors
JB: How much are people willing to pay?  Then how do you provide that to them and who provides that to them?  R: Pick a crop model (almonds, lettuce, etc.).  P:  I think it will depend on the economic value... P: You get down to the individual ... but who's going to pay for it ... when we were talking about yield monitors, I wish they'd give them away for free, because I can show the grower a spot that's hot and where his yield was down significantly. They can show the grower specifically and quantify for him ... the need for mitigation.  Until you get the yield data and show the amount of money that they're losing to problem areas, they don't see a need for these maps [from remote sensing].  The maps start to have meaning if you can quantify the loss of yield into dollars associated with certain parts of a field.  When I see growers make mistakes, it's because they're either not aware or don't think they can do anything about it.  I wish they had a yield monitor for every crop.  Then it would come down to, "Can somebody make a track for this variable rate fertilizer or chemicals?"  If you've got a sandy area, you can just add manure to that one area.  They don't want to have to add it to their entire 100 acres if only 10 acres needs it.  JB:  The indications that I've had are that  yield monitors are probably going to be available for several more crops in 1998.  P:  It's not just the straight yield monitors.  We've found that there's variation between field quality as well as quantity.

JB:  The cotton yield monitor that John Deere's working on may be ready in 1998.  There's  no question that yield maps will be a major force in bringing other technologies into use, because you're going to get a layer of information that'll show variations in yield.  UC Davis is working on the tomato yield monitor.  R:  If that happens, the next thing you want is research programs that's going to show farmers what to do.  The farmer wants some support on what to do.  JB:  I want to come back to something.  One of the things that's been driving in the Midwest -- they're finding that the variations between yield maps from different years don't correlate.  That's the difference I was trying to bring up in rain-fed vs. irrigated agriculture.

JB:  We've got a yield map that's going to help us, and the agronomist shows differences in the yield map from one year to the next, and now that's back to climate.  You can't control the input of water.  R:  You've got to normalize it.  R:  Yield maps shift -- your good spots one year may be bad spots the next year.  JB:  You can't pull the agronomics out of it by identifying high nitrogen levels in certain areas with variations in yield.  This is where the layers of information and the rain-fed agriculture is starting to throw up more questions than answers brought to the farmer.  I think it was unexpected, but I think we're going to continue to find more questions as the technology comes out. JB:  Wasn't there already a comment that extension personnel were overworked?  R:  I don't see that as a big problem.  That's part of extension's job.  Extension people will learn, just like everybody else.  That's an area that needs to be examined ... agronomy and economics. Also, our basic knowledge in farm management ...

Variability of Field Components
P: Right now, the farmers don't even know what they're putting down.  We've were surprised by the variability of every parameter.  The farmer was even more surprised than us.  JB:  There's been talk about putting different varieties into varying conditions so that yield would be enhanced.  I think yield is very important, because that's something that all growers understand.  If the variations in the field can be shown, then you try to address them.

P:  Just from the seed companies -- varietal selection.  They have 100 varieties out there and one irrigation regime.  So there are different varieties in the same field that have different water needs and different temperatures.  Remote sensing also allows you to make selections for breeding purposes too.  They get a higher yield with cool wheat down in Mexico.  You're breeding for biomass and [indirectly] for water use.  The problem for growers is they use the same irrigation regimes for different varieties and in some cases, for different crops.  I can show you things on hail damage where some varieties had 25% damage and other varieties had 60% damage.  When you get to insurance concerns, maybe they should insure one variety over another.  All this stuff I've seen, remote sensing can help.  We need some centralized group with deep pockets to coordinate it.  We have all the scientists, the science, the equipment.

Documentation of Case Studies
JB: Something I've heard quite often about -- why there's a small percentage of people looking at remote sensing is the average age of  growers is in the mid to late 50s.  There's certainly a percentage of people who are going to want to figure out if they can retire before they have to [switch to the new technology.]  The people who are computer literate and are looking for this technology maybe the ones who are the innovators leading the pack and showing the usefulness of the applications.  Part of the limitations on this, is we don't have a lot of documented case studies to show that this application will help.  If the farmers want a competitive edge, they don't want everyone coming up behind them adapting the technology.  Back to on-farm research, we should move some of the resources into some case studies and try to document the success.  Have some of the researchers come in from the professional societies and bring in a level of professionalism to where it's not biased and people wonder if it's for real or not.

General Recommendations
JB: If we look at the end users and their needs in general, how can NASA try to fill any of those needs or channel efforts into training or helping the learning process of the people involved?  The farmers can't be expected to be at the bottom level learning the technology.  They don't need to know the mechanics of the satellites.  If a grower wants to use information to help them in pest management, how can NASA help in that?  Is there anything to help train the extension personnel or people that transfer the technology into the private sector?  Through the Institute for Technology Development, we've seen some technology developed in the public sector brought across into the private sector and now the private sector is trying to get it to the people.  The end user needs to know how to take the information and use it.  If you take the information from remote sensing and put it into a decision support system -- usefulness of products, that's what the farmer wants.  Does he need expertise in helping him make that decision?

NASA should identify the technologies to the current professionals with expertise in agronomy, crops and soils.
JB:  One of the key areas I've seen -- I'm a certified professional agronomist -- I have continuing education requirements through the American Society of Agronomy that I have to keep my certification.  I'm a licensed pest control advisor.  There are thousands of advisors that have to fulfill education requirement.  If NASA could identify the technologies to the people who are in place with expertise in agronomy, crops and soils, then maybe the people there could be the effective bridge from science and technology to farming.  There needs to be an uplifting of the understanding at the professional level through existing professional people that are in place right now.  You have all these means all over the country -- the National Association of Independent Crop Consultants, California Agricultural Production Consultants Association and numerous other groups.

NASA should promote the system and set the standards for agricultural software.
P:   NASA could start thinking about an alliance with the private sector, so people can basically know about the technologies...  If you're a private company, you could basically take some of this information and try to enhance the value .  JB:  Some of the software that's out there now to analyze yield maps or to look at the different layers of information.  I'm not sure those opportunities are going to be there 3-4 years from now.  R:  The biggest problem you're going to run into is what's the private domain -- I wouldn't tell you how I do everything and give you the same software, because I spend lots of money ... Also your community needs $30,000 for software training on  a PC.  The software is not something where you can have a [one-time] 8-hour training period be effective. On ArcView or ArcInfo.  You can give somebody an image with keys to the software.  R:  They're not advocating that you take on that responsibility... R:  There's a group that's going to establish a set of standards for agricultural software -- it's the agriculture electronics.  What you're saying is the recommendation is NASA should try and promote the system and help them get the standards up. That requires that there be an open standard -- everyone has access to it.  P:  In San Francisco, NASA announced that they were trying to get away from data collection.  I think NASA's getting out of providing stuff for free to try and get more ... information from other sources.

Devote research and development into showing practical applications, pilot projects and case studies of remote sensing and solving anticipated and unanticipated problems.
JB:  If there's recommendations for the private or public sector, I think that would be beneficial.  Is there specific needs that can be identified on the training or the education or where to funnel research and development?   P:  I don't think the culture's the problem, I would say that outside of the state of AZ and CA, I don't know of a single real practical application of remote sensing technology at the farm level that we can implement.  I think of the recommendations... somebody's got to make it a self-serving thing.  You've got to do the agronomy and the research to develop... practices that we can extend.  I've heard all about this NDVI and I know about SAVI.  I've never seen or heard of one field that ever scheduled or did anything on a farm basis with any of that stuff.  Yet people are getting ready to put up billions of dollars worth of satellite that we can calculate NDVI and I've never yet [seen it in application]. R:  The laser was first introduced, people said it was a solution in search of a problem.  I think that's exactly the same situation with this technology.  The laser saw plenty of problems that no one could have anticipated, and I suspect that this technology will too.  But somebody's got to go out and solve those problems, and I'm here to tell you that research hasn't been done yet.  JB: To me I see a recommendation to try to do whatever it takes to facilitate on-farm research where we're not looking at small sites, whether it's development of case studies.  R:  There's a difference between natural history and ...science.  If this is going to be implemented on a broad scale, then people have to look at it in a scientific way and understand it. R:  I would say the majority of it will be driven by trial by error... I think that remote sensing in the same way.  If you know enough science to get by, I think most of it is technology driven, not research driven.  I don't think NASA will be able to tell me what's going on ...  I don't think that they're advanced enough.  R:  You're saying that precision farming is going to be trial by error, but they've got GIS, yield monitors, remote sensing data, statistical analyses...  Proposal for 1998 to do a study for remote sensing... If you look at what precision farming has done, it's trial by error.

Educate and expose people to the new technology.
P:... facilitate some basic education ... increase the exposure of these people ...  JB:  My point to try to bring that in is education, existing professionals, people who are going to be trained at the bachelor level now are going to be identified as the ones to replace the aging farmer.  Then professional societies could be identified to help facilitate and maybe key in  where to put the research and development.    R: Inter-disciplinary.  JB:  That's the key.  Inter-disciplinary -- background in environmental sciences included.

Involve Space Grant and university curricula in remote sensing.
P:  Space Grant could  help...and which has the money to do three things:  education, research and extension.  Up until now, space grant has [mainly focused on] education.  Remote sensing ... that's something that can be done nationwide at the undergraduate level.  JB:  Incorporate existing professionals or farmers, ranchers, producers?  R:  That would be the extension part.  At my meeting with space grant and NASA next week, I'd like to take some of your recommendations and [endorse and encourage] them to get involved with remote sensing.  The advantage is ... an organization that can try some new curricula here and there, and find which one will work better at the national level.  JB:  The universities need to evaluate their curricula to make sure they are training the next generation of professionals adequately in remote sensing.  Will they be able to manipulate maps?  R:  There are 500+ universities all over the country that are space grant institutions.  R:   I agree with education.  Can we really at this point in time know enough  about images?  JB:  There's a lot of ground-truthing that still needs to be done.

Devote an agency to focus on how to get customers receptive to remote sensing technology.
P:  It seems like over all the years, we keep focusing on this automatic machine that produces interpreted images.  That's not realistic.  Research is a wonderful thing, but there's so much information out there today that nobody's using.  It's not being used because nobody knows it's there.  If the consultants had it in their hands, they wouldn't know what to do with it, how to interpret it.  That's something that some agency, whether it's space grant or NASA or a commercial company, needs to focus on how we get the customer to be receptive.  We can't just tell NASA to do more research or send up more satellites.  It's time not to make the engineers happy, but to make the agronomists happy.

Demonstrate the usefulness of the technology for short-term and long-term management.
P:  One recommendation would be to demonstrate the technology to the end users.  The big companies are doing the technology, but they don't have the resources to spend money on demonstration of the utility of the existing technology.  JB:  I really think there's the need to demonstrate the possibilities.  I work with farmers and  I've seen innovations at the farm level -- when they have a breakdown, they find something to fix it.  If there isn't a part that can replace it, they develop it.  I think the farmer and the consultants can use this technology, but they have to have it demonstrated or at least brought to them what areas [remote sensing] can help them with -- pest management, plant stress, irrigation, mapping of soil borne pathogens, nematodes, drainage patterns, long term management, etc.  That just hasn't been brought in at the farm level.  There are two levels that I see -- short term management and long-term management.  A lot of farmers are more concerned with short-term management.  You try to show the value of [long term management], but the farmers are mostly interested in things that will help this season.  They can map year after years over generations. [Muffled responses] JB:  My point is that I've used the technology and know it's valuable.  We're not going to let people come in and make [teaching and demonstration] videos of our farm.  We're going to use it as a competitive advantage.  We're on the treadmill and setting the pace that no one else wants to get on.  One of the reasons we do become active and act as a strong player in the industry, is because we want to use the technology and want to see the satellites launched.  I know they're not going to do it just because one user finds value .  Somewhere the value has to be established, who's going to be the players and how to present it.  Then try to channel efforts into getting a lot of people to use it.  P:  [Referring to Recommendation #6] This is one of the recommendations to Space Grant by cooperative extension -- last year on how to bring remote sensing to the people.  Space Grant can make the video and distribute it.  Get education money from NASA, Mission to Planet Earth.  The other recommendation [#3] was to have some pilot projects involving real farmers and precision farming -- that's one of our projects in the southeast that will include media exposure, so that farmers will hear about it.  R:  In addition to education and government help, you can see the market developing slowly.  One example is Resource21 which started to sell their product 5 years ago.  [They've had limited success].  There is a lesson there -- the farmers need to be shown what they can do and then they'll pay for it.
 
All agricultural business majors should be educated in agronomy science.
JB:  One comment along those lines.  We have a big difference here in plant or applied science people in universities against people who are looking at ag business.  The business people may miss a lot of science and technology coursework and yet they're going to move into the management positions, making decision without the knowledge from  the agronomic side.  That's a real dilemma.  Another recommendation might be that all ag business majors study the agronomic side also.

JB:  We talked earlier about price and how much  people are going to pay for this.  I think there's a different between regions of how much people will be willing and able to pay.  High value fruits and vegetable in CA are going to pay more for information compared to the growers in the Midwest.  R:  That logic leads to the value of this data to Midwest crops [soybean, corn] that it is not worth very much...  JB:  So this is where you see remote sensing having value in crop yield analysis.  R:  That has value tomorrow.  I can sell that data tomorrow...

JB:  Do you want to expand on hyperspectral?  Will that fill a lot of needs and can we identify a lot of users?  R: ... wedge-shaped hyperspectral ... R:  Nothing is  currently available .

Tie in remote sensing applications to yield monitors.
P:  I want to raise another issue regarding yield monitors.  If you look at the data, quality varies across the field in a non-correlated way that has the same kind of variability that yield does.  So you have to look at yield, moisture, protein, carbohydrates, etc. as different parameters.  JB:  There are on-the-go sensors being developed to look at organic matter in the soil, read soil pH, looking for nitrogen levels.  They're looking at the possibility of calcium and magnesium sensors.  The level of research going into sensors is really advancing fast.  There's going to be a lot of layers that will tie back into the yield monitor map.  That's the value, because then you have something to go against all the other layers.  There are sensors being developed for plant population estimates, some that can be hand-held for field scouts.  So there's a lot of inputs and manageable factors that the farmers are looking at and that they can control on the seeding, etc.  We have a unit in CA that's going to be on-line in 1997 through Western Farm Service that will have three tanks on the back and will give you variable rate on nitrogen, potassium, phosphorus and zinc.  R:  That's operational in Iowa.  JB:  Yes, a lot of the development in Iowa and Minnesota is several years ahead of CA in variable rate technology.  I think remote sensing has an important role where we can start to see things during the growing season and go across time.  The sugar beet industry is looking at correlations between nitrogen and sugar development.  The nitrate monitors that they can get a map out of -- they'll go back with the yield maps and sample soil nitrates.  So there are people looking at different commodities, and what level of analysis is going to be different.  My point was that we're going to see innovations on the farm level that people don't know about.  We may just be thinking of the tip of the iceberg in terms of applications or trying to meet the user's needs when those needs haven't even been identified.

Develop a web site to link potential users to remote sensing companies and their products and allow farms to respond with their own application ideas and innovations.
P:  Maybe if we did something like a Web site where people can relate and see how people use remote sensing.   JB:  Can we make a recommendation that the University of California - Berkeley puts up a web site?  I see users tapping into the Internet - use is growing every day.  Is that something that the companies that are providers of the technology should be looking at or is that something NASA should be doing to identify the technology?  R:  Our purpose for doing it is to develop ideas ... if someone has a new application idea... JB:  That's the kind of thing I see farmers and consultants doing in the winter months when they have time.  They won't access that during the season when they are in production.  If they get into focus and get direction on where to go, no one wants to sit there going from address to address for leads.  It's got to get out to the farmers and the consultants.  I see consultants playing a major role as advisors.  They're going to bring the recommendations to the farmers -- precise recommendations from a specific technology and a specific format.  You don't want to be general -- you want to show how remote sensing can specifically help an individual farmer in his field.  When I presented remote sensing to the company I work for, I sold it just on pest management.  White flies were a major issue.  Then we found irrigation management to be a big bonus.  We weren't even looking for that.  That was my point earlier, that we don't even know what we're going to see [in terms of needs].  So specifically lay out what the TM will do, airborne systems, etc.  Until that happens, we're only going to have 5% of the end users using the technology.  The bottom line is to get people using the technology, making money so that the satellites will be launched.

Ground-truth applications of remote sensing.
P:  Can I change the scenario a bit?  If you don't provide the information and ground-truth the knowledge out there... JB: You mention a good point, because you've tied in information with ground-truthing. The grower doesn't know how to ground-truth or doesn't know what to do to evaluate it.  We're not at the point where we can identify nematodes on cotton, etc.  We don't have enough signatures identified.  Ground-truthing is a very important part of this.    R:  You're not talking about a whole section of production -- you're talking about a specific ag production model and very specific section of the ag economy... JB:  The point is whether you can be more effective with your time and pinpoint where they scout.  If the scout comes back and says it's spider mites, the pest control advisor writing the recommendation is going to verify the need for miticide and not aphids, instead.