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

October 23-25th, 1996 
 

Session 2:  What Remote Sensing Capabilities Will be Available for Agricultural Research and Commercial Applications in the Next Ten Years?

Transcription Notes
 
| Ag 21 Agenda |
 
Moderated by Jack Estes
Session 2;  October 23, 1996
Allewelt Room, Alumni Center
Note:  Interjections are in [brackets].
JE=Jack Estes; P=Participant; R=Response
“...”= Words muffled, cut off or interrupted by other participants.
 
Introduction
1. What remote sensing applications will be available for agriculture research and commercial applications in the next ten years?
2. What platforms and sensors should we focus on at this time in agricultural applications?

JE:  I’m a geographer from the University of California, Santa Barbara (UCSB).  There is such a wide diversity of potential applications in agricuture remote sensing -- from the broad, low resolution continental scale, (e.g. Using active microwaves or satellite passive microwave devices to look at starting soil moisture conditions as input to photography for crop growth models before the crop is even in the field) to high resolution fine scale (looking at diseases in s fields, such as a blight).  These topics are not easily subdivided.  Sir Alfred J. Mckinder, a geographer who was a developer of the heartland theory, said,  “All knowledge is one.  Its division into subject matter is a concession to human weakness.”  It’s hard to break this down in an effective way, but that’s what we’re asked to do.  I’ve been in the field of remote sensing for about 33 years, and I am still learning.  Hopefully today we can all continue to learn together.

[Everyone introduces themselves:  Gary Shelton, Office of Mission to Planet Earth --working to put the NRA together, background in airborne remote sensing; Dan Lustian, Agricultural Imaging Exchange; Shrini [?], Bio and Agricultural Engineering at UCD -- interest in precision agriculture; Dan Barbar, Dalencko[?] Who are a purveyor of pesticides -- how remote sensing and precision agriculture technology will impact/shape the future; Carl Schueler, Hughes looking at next phase in his business, which is to expand contracts for instrumentation and customers; John Syder, one of Susan’s [?]--worked at Aerojet, Researched sensor design/applications; John Ransom, NASA Goddard -- been working in NASA for 20 years, representative of NASA branch, strong interest in agriculture; John Grace, Resource 21; Ed Sills, grower in California, agronomist with the American Farm Bureau Information Technology Committee; Lon Crosby, operates technology company in Iowa, farm on the side, reporting for Grant Mangold, Agriculture technology newsletter; Bill Stoney, Mitretek -- small contract with NASA to answer Congress’ questions about LANDSAT 8.  {See Following Discussion}]

[Bill Stoney]
LANDSAT 8 is what follows LANDSAT 7.  Congress put in the bill that they had to have the answer to 4 questions about LANDSAT by the end of next year:
1. Should NASA fund another LANDSAT satellite?  We can make the continuity argument.
2. Or should they buy the data from the private sector?  And the “should” is, can they?
3. Or should they be part of an international consortium?
4. Or [should they be] some combination of the above?

This process is interesting to me because agriculture is “a potential user” of the data, and this decision might be important to the National Aeronautic and Space Administration (NASA) in one way or another.  There’s lot of pressure on NASA to get out of the LANDSAT business.  We’ve written a preliminary report stating that the private sector really is going to enter into this area and will to some degree compete with LANDSAT pretty strongly.  Foreigner nations are already in it.  You can buy equivalent LANDSAT data from a number of countries.  Some people say the data are not equivalent, that they are not as well calibrated etc., But those arguments are a little bit specious.

The other recommendation made in our report was that NASA look to its strength particularly in the area of the development of new technology.  Obviously, NASA is not the leader anymore in optical technology.  The major point here is that there’s no place doing research on the sensors themselves, that are really going to be applied to these new and different things we are asking for here.  Something beyond what you can get out of the current sensors.  You mentioned soil moisture --pretty important for some applications.  We don’t have a sensor that would do that at the scale you’re talking about, but the technology is there in some form. I mentioned radar this morning.  The U.S. has fallen behind in radar, and we don’t really have the background and plans to apply it to civilian uses.  That’s an option that NASA should be looking at.  NASA should also be looking at multi-temporal thermal.  This should happen in 2004.  NASA typically needs lead time so we’re really talking about something that needs to happen by 1999 when they’re going to start doing something seriously and go after money for it.  I have been in the LANDSAT business for a long time in NASA, and I know how they work.  To continue to upgrade our technological capabilities, we must begin to act now.

What remote sensing satellite capabilities will be available in the next ten years?
JE: The best place to start is with the charts that Bill [Stoney] used this morning on satellite technology.  What can we add?  What on there isn’t right?  What would be the priority order? If you remember, we’ve learned that there are going to be 3-5 satellites in the space imaging suite.  They’ve got two in the barn and so on.  MOMS is not up yet, ADEOS is going a little earlier, LANDSAT 5 will still be operating into the year 2000 and so on.  I’d like to have a consensus that we looked at this list, we accepted this list as complete and I’d like to move onto aircraft or others. [Muffled conversation about LIGHTSAR] [R=Response] There’s no way [LIGHTSAR is] going to be up before 2000, but sometime in the 2000+ time frame.  The only thing I’ve heard about LIGHTSAR is that it’s going to be developed by the Jet Propulsion Laboratory (JPL) with commercial partners. R: They had a meeting at Sioux Falls about LIGHTSAR and JPL and commercial partners. NASA tried to make the case to the commercial developers that it was a good deal.  The developers said there is no business or market, and they wouldn’t put up any money until there was a market.  NASA is wisely going ahead and trying to get research money for it and put it  up. R: The whistles and bells will depend on how much money somebody else will contribute to the effort.  NASA’s going to have a fixed budget for the foreseeable future and if you want to do other things, then commercial has to come in.

JE: LIGHTSAR is a potential addition.  Anyone know of any thermal systems that may go up?
R: I know of a wanna-be.  The people at JPL are also the developers.  The system is named after the Indian guide to Lewis and Clark, Sackajawea.  This sensor is not actually funded, but it’s certainly something that’s has been put up and will come up as a candidate for the next round.  It’s multi-channel thermal.

What remote sensing aircraft capabilities will be available in the next ten years?
JE: I look at aircraft as an important aspect of any NRA or any research program that NASA is doing.  R:  What’s GER doing?  R:  Last I heard they had venture capital which last them the next couple of months and were going to try to do something with [?] And Lockheed. R: It was kind of a smoke and mirrors type scenario in order to get one [company] to release money first to stimulate some funding strategy that they got.  As far as I know, Lockheed isn’t doing anything with them right now, TRW [muffled] and that was it.  That’s all I know about GER.  R:  Lockheed has come to our campus recently.  [Muffled]  The way I understood it they are putting in four bands [muffled] ASYMETIC[?] By December 1997, and they are predicting an increase to 16 bands, so they could potentially be blocking that information, and they are trying to find out whether it is possible to do that.  There’s a lot of money [muffled], so Lockheed was really doing some leg work to find out what’s ... Or whether they should be ... They said they would be ... Their investment in this company would be 15%... R: I’d like to pass around this copy of Earth Observation Magazine where they made an announcement.  It doesn’t say anymore than you said.   Ten meters, high resolution band ... And they need $300 -400 million.

Need/Ability to Discuss Existing/Potential Aircraft Systems
JE: One problem I have is fairly easy because the satellite systems are the big sensor systems.  They’re visible.  Everyone can see them.  Should we even try to focus on aircraft systems?  I can’t define all of the available aircraft systems that are going to be operating.  I would make the point that any sensor system that is in development now that will be ready in the next ten years could be available for aircraft.  I mean we already have radar sensors, thermal sensors, passive microwave sensors, UV sensors, LIDAR on aircraft. How much do you want to pay?  I’m trying to bound chaos.  With respect to aircraft capability, I would also posit that if you’ve thought of it and it’s being developed by a firm, it could be there for this kind of research.  That goes across the electromagnetic spectrum.  I’m not saying we don’t need [aircraft systems] – I just can’t think of how I can go out and name them all.  R: Nobody can.  There is too much going on. R: I think government operations we can identify, but commercial operations, I wouldn’t have any idea. R:  There are so many “Mom and Pop” operations.

Potential Research/Products of Remote Sensing
Use of Remote Sensing for Insurance Firms
P:  We looked at this as a problem in Thailand. This goes back to a meeting of the Board of Directors of Rain and Hail Insurance Services.  I have an agribusiness firm and we’re looking at the possibility of using remote sensing in the area of loss adjustment, acreage identification, etc.  At some point with the ability and the cost, there’s an increase in cost and a decline in the value that you receive back from the crops that you put in.  It’s kind of like squeezing an orange.  At some point, you’re squeezing and [barely] getting one drop of juice out of it.  Whether or not that additional drop of juice is going to get you any more than what you have or what you’re striving for becomes the concern.  We really had two objectives in pursuing this.  We wanted to see certain products that could be given to the growers that the growers would use.  It seemed to me that the general product having to do something will take a little bit of time, because it varies by crop and by grower.  It seems that it was going to take a good deal of time (3-6 years) to truly get the grower adapted community fully integrated by using remote sensing.  You have some big players that we can list here in California, or around the country that will come on board immediately.  Whether or not they get any value out of it, they’ll come on board and actually do something with it or at least take a look at it.

Moisture Content
P: Moisture gets us into irrigation scheduling and water management.  In trying to find another way to look at the problems on the field without having to write some type of biomass model or vegetation model, which in many cases you can’t do and just can’t get the information you want out of these models.  We need to find another way to get the moisture other than writing some type of vegetation model.  You can’t get the information out of the bands that most systems are using now.

Biomass
P: The biomass issue: Isn’t that a function of bands? R: Not necessarily.  From what I understand, out of the 4 bands they are using and whatever additional value they may want to get out of a few more bands of a hyperspectral system, they should be able to do both moisture and vegetation, biomass, planted acreage and some of the other things they are looking at.

Chemical Analysis of the soil
P:  It’s not so much the chemical analysis itself, but users want to see if they can identify a problem area in the soil and then get some kind of soil map and calibrate it based on ground-truth findings and whatever else it would take to do that.  You have to be cautious, because the growers are not going to give up their scouting systems.  They will never get away from using people in the field.  We may help to reduce the number of people they use in the field and making them more effective in what they do with calibrations -- a percentage on a given location will be much more accurate that what we’re currently getting in many cases.  So, to think that you have to be able to do all of that from a space [satellite] or airborne system, it’s overkill, because you’re assuming that you’re going to totally eliminate all the field inputs when you’re not.  The grower will never get away and fully trust the remote sensing system.

Radar
P:  I heard the radar issue coming up -- wanting to see through the clouds for damage, flooding, planted acreage, etc., Or maybe they felt that they just could not assure that they could get data otherwise --  I didn’t find that that was necessarily an issue.  With radar, some cotton people said they’d like to know some things about root structure, but they couldn’t say that they were really concerned with or how much would they pay for that information.  With radar, I’m not sure they would pay for a round analysis of root structure for cotton at the cost it would take to put up a satellite to do specifically that.  Getting to the grower -- providing the grower with a fully integrated GIS information package that he can choose to use a portion of or all of is one thing that  I think it will take a fair amount of time to get to.

Planted acreage, crop identification, and damage identification.
P:  Another big issue that agribusiness can help carry this whole industry is providing users with planted acreage, some crop identification, identify that damage has occurred, not necessarily what the damage is.  Resolution -- Space Imaging, can get 1-4 meters depending on how they’re going to apply it.  That’s more than enough.  Resource 21 will tell you that you can get the planted acreage, and crop identification (ID) at 10 meters.  Most of what you have is already here.  The next thing is combining the resources. If I were NASA and I wanted to do something that would really have an impact, I would look at a series of inexpensive satellites to look at the broader picture.  That would include planted acreage and crop ID.  If they are going to get into a real high resolution satellite that cost $500 million to look at root structure, I don’t think they’ll ever pay for it.

SARs and Lasers
P:  I think we might want to come to this issue of smog[?]. R: We’ve got interferometric SARs, and you can do interferometric SARs from aircraft. R: We need have some active optical sensors. R: Active optical? R: And that’s laser systems.  With LIDAR, people are thinking “well what if they were pointing those things at the ground and not only the topography of the surface but the penetration of the laser through canopy ...  JE:  You’re right.  There’s passive and active programs at Goddard that do that. R: That buys you the capability to start looking not only at reflectance but fluorescence, and at one level, absorbence. R: It’s being used in UV right now - GIS ... Simulations. R:  And if you want to sit and predict what’s happening, while we are sitting here, Researchers are combining genes that are producing light as the mechanism track the genetic engineering products. R: You mean proprietary planning? R: You’re planting so many acres of round-up ready beans and here is the bill for your genetic engineering.  I don’t care whether you buy it from me or buy it from somebody else. R: Is this something you want to do to track it? R: It becomes real easy, I mean it’s a piece of cake. R: To do it because of the letter? R: Well, once you have certain kinds of imaging capability and you can trace it in a geo-referenced procedure, you can tell exactly who is planting what, where and what kind of bioengineered genes they used in those plants. You want to know that because you want your royalty payments. R: And if you were into natural products, you may want to be able to prove that you didn’t do something. [Agreement]

Microwave Use in Remote Sensing
JE:  Is there anything else?  Some of these type of products, at least in the near term, are only going to be available from government sources.  Some are available widely through commercial sources. [R muffled] Some of the microwave companies are looking at having smallsats in two years ... and it has to do with the soil moisture ... R: I remember the interesting thing we had with Earthstar at NASA, we wanted to do something with a C-band or an X-band and combine it with an L-band and looking at the curve, try and get a soil-moisture budget based on some of these penetrations...  Microwave is one of the areas that NASA has never really been able to develop R:  We have SIR-C. R: But if you’re talking about agricultural applications, soil moisture become important, not only from a planning perspective, but from an insurance perspective, because whether you did a timely operation is going to affect the charge an insurance company should make for providing crop insurance. R: I suppose this covers all the basics … the kind of sensing algorithms ... The imagery looks to be transition based so they’re looking at the atmosphere ...microwaves ...

The Effects of Pollution/Need for Atmospheric Correction in Remote Sensing
JE:  We’re just saying what will be available in the next century.  The real issue that I’d like to get to after this chart.  If we were going to do research for agriculture, where do you want to put money? R: I think another area we’ve talked about here ... We can also lead to .. Agriculture in terms of ... Airborne ... In the troposphere ...  Most of the birds that are up there now are ... A lot of content in the troposphere -- sulfate particles ... R: And how does that relate?  Weather? R: We’ve got some way of tracking this in agriculture, so that then when you get a bad airplane ride down the road, where is it coming from?  [Why isn’t it in] focus here?  Is it solid particle or sulfate particle?  ... Right now we’re ... Particles and everybody shot us down and these guys suffered.  Between here and the South Coast ... We’re going to go across the rest of the U.S. because right now health issues, health pollution for safety[?] Anything less than 2 microns ... And this is the next battle zone.  We can do it with aircraft, but we’re talking about where are the researchers? ...

JE: Should we take note of  Geostationary? R: Geostationary.  There’s two things you’ve got to look at.  One is the kilometer and above systems which includes the MODIS.  The other aspect is atmospheric correction.  We do not have any systems for atmospheric correction in the system, except currently Landsat 5, but MODIS will have it.  We don’t have any low resolution or high resolution guys that have that as a targeting.  That’s one of the issues when you get back down to the higher resolution stuff that could be a real swinger.  We’ve got some insights into it lately.  I don’t know whether this 250 meter resolution thing they’re going to put on EOS is any good.  They’re going to put a 250 meter wedge filter 4 band system on PM-1.  That’s the pixel size.  They’re going to do a gross one and its a little add-on instrument that’s going to fly right along side it with the idea that they’ll develop this and sell it to Resource 21 or anybody else.  Because they did put a cloud indicator filter system sensor on there. R:  For this NRA, the time frame we’re talking about, almost everything up there is available for people to use. R: What does “available” mean? R: If somebody proposed this on the NRA, I think we could supply that. R: What does that mean in terms of the research agenda for the private sector and the government sector? R: We could add that question.  That’s what I’m getting at -- I think we’re defining some of the questions by saying that, “Agriculture needs this. This toy is around.  We might want to propose how to make it.”

JE:  There’s a second issue. Just to kick around some ideas and give NASA some input:  Where does this group see where there’s high potential for agriculture or commercial applications that aren’t being used now?  A lot of people forget, there is a lot of verifiable firms out there that provide remotely sensed data commercially. R: How about thermal, anyone really doing that?  P:  No, but NASA’s got that too.  They’ve got a four band thermal thing down at Stennis. R: It’s not at Stennis.  It broke Tuesday night.  Badly broke. R: Well, I can’t keep up with that. [Laughs]

Application of Weather Information to Agriculture
JE:  LIDAR? R: I haven’t seen LIDAR’s looked at from the ground. R:  Primarily they’re thinking of using for this for the wind over the ocean and the wind of the troposphere. R: What’s the weather information do?  Once you add weather, and say weather is an agricultural application, then we have to add in all the information. R: The most important parameter in agricultural applications is weather. R: Should we put GOES on with these others?  Should we put more of the weather statistics on that? R: We could.

A Systems Approach to Remote Sensing
P: There are three things to evaluate a sensor, from my limited experience.
What kind of knowledge is needed?
I’ll start with a simple story.  Three years ago, people started looking at site specific technology. The first thing they looked at was they tested the fundamental theory of agriculture in the Midwest: There is a very strong relationship between the type of soil that the crop is grown on and the productivity of that soil. If you pick up an agriculture book that’s what it says.  It is so strongly held as a belief that the tax structure in Iowa is defined by soil type.  Yet the people who look at the relationship between soil type and yield show no relationship.  They called in the USDA scientists to come along and remap the fields, reran the analysis and still found no relationship.  They thought that it must be the deep structure soil type that’s important.  They drilled holes 30 feet deep looking at the subsurface soil type to try and find the relationship.  No relationship appeared.  We now have multiple years of research that confirm the lack of relationship between soil type and productivity, and it has now been repeated across the Midwest.  If the fundamental basis upon which all textbooks are based is wrong, what is known?  In you’re thinking about sensors, how do we establish the fundamental knowledge that drives agriculture?  And how do we assemble a set of sensors to do that in a cost-effective way?

You have to look at sensors using a systems approach, rather than a platform approach.
When you’re thinking about agriculture in knowledge needs, you can’t stop there.  You need to marry that with a comparable system from aircraft.  If you don’t marry the aircraft systems back with systems that can be used by producers out in the fields as they go out to walk, to physically confirm what you’re saying is a problem, you really don’t have a totally integrated system.

Think about what agriculture is physically going to look like five years from now.
Production agriculture in this country is changing fundamentally whether anyone wants to admits it or not.  The family farms with 2-300 acres don’t exist anymore for all practical purposes.  The producers you are talking about are, by definition, the 200,000 farmers in this country who are going to produce 95% of the produce in the year 2001.  That’s much different than using USDA definition that there are 2 million producers out there in this world.  If you don’t turn around and say, “Here’s the knowledge I need, the sensors systems approach, and the people that are going to be using it, you can’t make any objective evaluations about the kinds of systems you need to develop and implement.”  So much for my speech.

Expectations of Remote Sensing and Market Forces
Current and Future State of Farming in the U.S.
R:  It’s a good speech, because it really gets to the heart of what they’re asking us to do and they’re trying to find out what they should do for a study program.  That certainly sounds like you have defined one.  The element I would pull out is that NASA ... Look at the changing environment [of the future].  That ought to be an element of it.  What are farms really going to look like and who are we dealing with?  Is it the owners or the workers? R: The point there it’s not all a homogenous thing either.  There is the demise of the family farm.  The family farm in the United States has been a myth for a long time, but you’ve got a series of different kinds of farm operations all the way from the big producers from the San Joaquin (corporate agriculture) to a guy who is farming 5 or 6 leased lots in the Midwest, to some people who are still farming the “family farm”. R: But, the “farm” is now defined as 2 or 3,000 acres as opposed to 160 acres.  And few, if any, of the farmers in the Midwest owns all 2,000-3,000 acres that he or she farms.
P: You’d better change your perception of what farming is going to look like in the next five years.  Because there is a massive change going on. R: Listen to his predictions.  There are only going to be 200 farmers in the country. R: Well, 200,000 farmers in the year 2001 are going to be providing 85% of all the produce. R: California still produces 254 different kinds of crops.  That’s not 85%.  You’re correct that 200,000 farmers are going to be providing 85% of the produce, but not the rest of the high value commodities. R: Okay, how many producers of celery in CA today.  How many will there be five years from now? JE:  You guys are beginning to ask rhetorical questions that are beyond the scope of what we’re trying to get at here.

P: What about the DOD data that’s becoming available -- I realize a lot of that is older data, but what about that as a future data source? R: Bill partially covered that with that WARFIGHTER satellite story. R: Timeliness is the essence of agriculture and whatever system they come up to use that data, you’re not going to get at that faster.  JE:  I don’t know, but the DOD needs data awful fast too. R: The declassification process is going to stop anything else. R: I’m not sure I want to help the DOD make use of their satellites for agriculture in this country.  Besides, I think they’re not appropriate.  They’re by and large black and white, and they are high resolution and they’re not into the, except for the WARFIGHTER, where they’re going hyperspectral.  That’ll be available through a reputable source. R: The area coverage on that is going to be small.  We’re talking about a couple of nautical miles square. R: No, wrong. R: The next generation WARFIGHTER hasn’t been designed yet.

JE: If we had to focus on upgrading existing technology that would help agriculture -- I was told that thermal infrared, microwave, lasers are the places where NASA should focus at this time.  Data through-puts are a major issue and calibration is also a major issue.  Those are the issues in this area.  We’ve gone beyond our charter.

Expectations
P:  I’d like to talk about what information can we expect to get from remote sensing.  A lot of the information on agriculture that has just been acquired since GPS type systems have got going, instead of replicated projects and reversing directions.  What I’ve heard is that things are not as it seems or would be expected.  We thought that we could just look at nitrogen from the color of the leaves or things like that and predict yield.  From what I hear from the people who are fixing and researching the yield monitors, that’s not happening and there are other things that are going on that we don’t know about yet.  If there are other things that we don’t know about, how do we know what to look for?  JE:  That’s a separate question.  And the way that’s a separate question is there is a whole suite of sensors and breakdown of spectrum in a variety of ways that you can have.  Like Bill laid out, there’s is no magic.  You pick spatial resolution, spectral resolution, and temporal resolution and you try to maximize an object to background contrast ratio.  That’s all you’re doing with remote sensing fundamentally.  You’re trying to see a difference either in the soil or in the leaf structure or the canopy or a mix of all of the above. R: Or the air.  JE:  I am not an advocate that remote sensing can do everything.  I know some things it can do, and I know some things it can’t do.  I know some places where it has been oversold and I know some places where it has been undersold.  But what you’re asking is a research question. R: No, just suppose that you find out that the thing that determines yield most is the activity of soil bacteria of a certain type.  Now suppose that these organisms evolve CO2 more rapidly.  Suppose they do something in the soil as far as changing the color of the soil -- this is all hypothetical. R:  No, it isn’t unfortunately. R: Don’t you need to know that before you design your sensor?  What you want to look for?  JE:  Actually, you’re absolutely right, but what’s happening because most sensors are what are called “facility systems”, and very few commercial vendors are developing application specific systems, they are trying to go for a broader market in developing facility systems.  They are not focusing on that one “killer” application unless you can prove to them that there is a commercial market for that and a profit motive.  A couple of examples:  LANDSAT was developed with a crossing time early in the day allegedly for agriculture to get away from problems in the Midwest where you have convective clouds in the afternoon and your clouds build up over time.  It was also at that early morning crossing time because of the low sun angle photography for geologists and geophysicists.  That’s what USGS wanted at the time.  So, you have USDA on the one hand and USGS on the other hand.  Plus, if you had every 18 days a repeat cycle in the earliest LANDSAT, you had the probability of catching two dates in the growing season in a given month in the agricultural belt in the Midwestern U.S.  That was another thing.  So, it was a compromise.  LANDSAT, at the time -- the old MSS, was a state of the art scanning system.  If we hadn’t thrown it 565 nautical miles up in space at the time (I think that was the orbit at the time), you could have had a different resolution.  Because it was 565 nm., the spatial resolution was 80 meters.  Everything is a trade-off here, and that’s why we have these hypercube technologies, and so on, to research the very kind of problem you are talking about right now – these high spectral band systems to talk about the very detailed types of answers users need.  Yes, we need to know those.  If somebody can define a problem that is so economically important, that it’s going to wipe out a major crop in the U.S. and people are willing to spend so much money, you’d have every company in this room design sensor technology to address that question, right now.  This stuff is all traditionally driven by forces that come out of the Department of Defense (DOD) and most of the remote sensing systems in the U.S. came out of technology that was originally developed for the military who spent a lot of money.  I don’t know any other way to say it. R: But you can still create a matrix that says, “These are the sensors that can be conceptually built and these are the capabilities that these sensors can provide.”  This is the information stream [Agreement by JE]  NASA, in concert with or in some combination of public and private opportunities, needs to define the mechanism where those systems are available at least in an airborne platform and can be used for objective agriculture research endeavors.  And then, you at least provide framework data for somebody to start making some decisions on.

P:  NASA has a program down at Stennis called the Earth Observations Commercial Applications (EOCAT) program, which any company can come in (a few companies have come in) and if they’ve got an idea for a remote sensed product using the satellites that are up or using their airplane, Stennis will (if they win the contest and they give 4 or 5 contracts a year) use their facilities to help teach them how to spell radiometry and give them samples of the data and try to get them to work and put together a product that the company defines. What NASA requires is that the company puts in an equivalent amount of manpower, time and effort (no money exchanges hands).  The idea is that NASA wants the company to be serious about getting their business going.  That’s something that is going on now and it is certainly applicable to the agricultural issue in defining product.  It should be looked at.

Verifying Remotely Sensed Data
P:  Here’s all the aircraft technology, all this theoretically could be moved into space on a satellite.  Where’s the problem in moving this technology down to ground level?  Because you have a plane that’s flying and has this kind of resolution.  Now you need to come back and “ground truth it”, you need to do some correlation.  So in the visible range, I can go out and buy a spectrometer system, that I can plug into a PC for a couple of hundred dollars that will give me a thousand and 24 bands that 5 years ago would have been considered the ultimate in research level resolutions.  That’s great because I can now scan my soil, I can scan my leaves, I can scan water or do all these things. If you don’t have that [ground-truth verified data], you don’t have a system. R: We use models a lot of times.  If you take radar out in the field and use a scatterometer, there’s a lot of work that needs to be done.  R: If you want to look at root structure of a plant growing in Georgia, someone is going to have to carry out a ground penetrating radar system out there and say, “Okay, here is a plant growing in this environment and now you have to systematically dig it and slice it.  It can be done, but who has that ground penetrating radar system to back up these things. R: Isn’t that an issue that the guy who wants to do the NRA will come in and say, “I’m going to do it, i’ve got the ground system, you’ve got that, we’ll put it together.” R: Well, if you’ve got these pieces, yes, somebody can do that.  It becomes a marketable opportunity. R: They’re going to give a couple million bucks away to do this stuff. R:  I would be willing to bet if somebody came in and said they want to do ... [muffled] would not be funded. R: The reason is this.  I have very seldom ever seen NASA fund something that had a significant accuracy assessment. R: Then you weren’t listening when they did the whole big program to find the accuracy of the radiometry of the TM. R: How much ground based investigation was conducted? R: I don’t know how they did it. R: I thought a lot of it involved evaluation of large targets planted into the White Sands New Mexico Test Range, Everything for both radiometric, spatial resolution evaluation. R:  There were two experiments that I am aware of. JE:  I’m not saying it cannot be done or would not be done, but to get down to that level of detail you would really have to convince NASA of a major payoff having very large economic benefits. R: USDA might act as a co-funder. R:  That would be a stretch ... R: We have to do two things.  One, we have to show that there is some feasibility of employing airborne and space radar to do something.  Then the next logical step is to try to operationalize that capability.

Need for a Report on Abilities/Accuracy of Existing Satellites
P:  I’m not sure you’d always necessarily have to match on the ground exactly this information, although that’s important.  What you’re really interested in is, what are the actual parameters that exist in that field at that time? R: You just brought up a question that I think we ought to think about.  NASA does not have any source that anybody can go to that’s interest in using satellites and find out what is the accuracy, what is the calibration, what does it really do?  There is no way you can find out whether it will do your job.  There’s no consumer report to help you choose which sensor system or set of sensor systems a user might want.  Part of the NRA, for this particular thing, could be to help evaluate for [people that come in and say they want to evaluate] crops for these agricultural applications.  Scientists blow them off.  That would be very valuable.  There hasn’t been any source of funding in NASA to do that and that would be a big event for the other sensors that are out there.  In this country, we’re simply stuck with, if we can’t get LANDSAT, we don’t want to use anything.

Legal Issues associated with Remote Sensing
P:  All you have to do is look at a very specific question.  For example, the Mississippi River basin - what kind of rules and regulations can you expect to come out of the Sierra Club’s lawsuit and its ramifications -- that’s a perfect remote imaging opportunity where, in essence, you’re going to get the entire heartland involved in remote imaging whether they want to or not. R: We talked about a lot of things -- but one of things we didn’t talk about was legal issues.  It’s going to be used for government surveillance of legislative compliance. R: Enforcement issues using all of the capabilities. R: It goes beyond that too, because I can fly fields and I can see that this herbicide didn’t work under these conditions.  Here’s some objective data that I can sue dupont or whoever for non-performance of their product.  Or I can now document that overspray on a crop had an impact on productivity and sue you.  This technology changes the whole approach that people use agriculture. R: Using technology appropriately, you can now prove the negative.  You can prove that you didn’t do something to a piece of land e.g. didn’t apply a herbicide.  That’s an important issue to the natural products. R: Should we put the genie back in the bottle? R: No!  He’s out, you’re going to have to deal with it. R: The other thing it does is it allows farmers to do more experimentally and test from even more on-site types of evaluations. R: It’s far worse to become the next generation of managers and “professional scientists”, if you will, and there’s no reason to have an extension service.  If you sit down and say, if producers can access knowledge and they can farm their farms in a way to generate knowledge, how long does it take them to create a knowledge base that is better than anything that a pioneer can feel.  It’s basically one year.  That changes a whole lot of political issues and relationships.  R: Most farmers that I’ve interacted with, I would say that a very small percentage of people are really going to interact with this data themselves.  Primarily they’re looking to farm coops or the people they normally deal with (crop consultants, service providers, fertilizer companies, etc.) To expand into the GIS, remote sensing capability and that would be another service that would be provided to them in terms of a final information product of a type that they would desire.

Cost of Remote Sensing
P: You can get about $8-10 per acre. R: Depending on the crop, realistically, if you use a fairly basic system like what Space Imaging is putting up now and what Resource 21 is looking into, you could get anywhere between $8-14/acre.  Now I’ve heard people says, “We would like to make, try to get enough swing on it, percentage of revenue increase, etc.” That pricing strategy would never work.  We would like to get it under $4/acre.  If you’re going to do row crops in the Midwest, where it is non-irrigated, $1/acre is a push...  R:  Does a grower really want to know planted acreage?  I assume [they already know that]. R: It’s not for the growers.  That would be for agribusiness.  JE:  That’s the problems we have with talking about a wide variety of customers for the outputs of image data.  If you’re going to go at one market, there’s always a problem with remote sensing ... The right one image.  A lot of people will look at a product or product suite if you process and produce the right kind of product that they need in a timely and efficient fashion.  R: They need [planted acreage] while planting is occurring before emergence.  There are strategies that you can use to generate that information. R: They would be thrilled to death right now just to know how many almond trees are in California, Period!  That’s not even in terms of acreage.  In many cases of talking with cotton people, they would like to know, once this plant comes out, “What’s out there?  What can we expect?”  Because right now, the swings in importing are so dramatic outside the U.S.  The market place is just going up and down, and they are actually taking both sides of a position.  That’s really a problem. JE:  So we’re back to the commodity end? R: When you talk about the bigger picture, when you look at what’s going to drag remote sensing out to the grower, assuming that the individual growers have a fully integrated farm system that costs a lot of money to develop, it is going to take some time to develop the systems, you’re going to have to find a way to give it to them as inexpensively as possible, be it $1, $2, or $3/acre, how are you going to do that?  You’re not going to be able to go out and put the whole development bill, if you will, on that 2% or 3% of the early adapter community.  They’re not going to carry that cost -- you’re not going to get $500 million out of those people quick enough to cost justify the system.  You can come up with the scenario that produces a satellite system or constellation that will do the broader-based stuff, the simpler stuff up front.  Then you’re able to push it through the system and you’ll have the time to get the curve of the growers using it along with agribusiness and hopefully they will combine to drive down the price.  That’s what our strategy is and that’s exactly what we’re trying to do.  R: The problem is here, that there is no one or “the” problem.  Everyone here is making a lot of generalizations.

JE:  There has been some discussion as to whether NASA is even going to or even should stay in this game.  Our opinion is that we see a lot of work going on with the high resolution and trying to resolve really finite points out there for the individual growers.  We feel that it will take forever and a day to get to that individual grower and get absolute market saturation and that’s going to be a real tough challenge.  Maybe now is not the time to develop that system.  But what will help get us there is maybe some of the more simpler systems that can do things that we know we already have the capability of doing.  Take a little bit slower approach to work with those issues.  I know Space Imaging is looking at, “What do we do for our third satellite?” Which is why we’re visiting with you.

P:  One thing we all ought to know is that LANDSAT, by law, is going to fly.  The data is available to anybody for the cost of reproduction.  That’s going to be the policy of the government by law.  Which means for entrepreneurs, if there is a market for using the large scale data, there will be an opportunity starting in 1998 to have a satellite coming back with cheap data, and anybody can make a business out of it.  No restrictions. R: As the AIE, that’s what we’re looking for.  If you have that available we would acquire it and then drive it back into the business community and then let them value enhance it, resell it or make it available to the grower in a way that the grower can digest it and utilize it.  We’re willing to do that as a business.  The simpler approach right now, we would like to have today because that’s sellable today.  The other things we think are it’s going to take time to develop that community.  But you’ll lead with the simple information first and follow with the more integrated systems.

Cost of Commercial Applications
P:  Has NASA done some work in what I call the minimum threshold requirements associated with a specific application? R: Not that I’ve heard of. R: That’s an area that hasn’t been done in the past, primarily because NASA’s objectives are to advance the state of the art of global sensing systems to find out more about a systems science, better sensors, as well as we can. Commercial applications require less expensive equipment than what NASA has flown.  The equipment was designed to be just good enough for the job, but not more expensive than necessary.  So the quality of the equipment is not the issue.  The issue is:  what’s the poorest spatial resolution that I need to use with this application?  What are the broadest spectral bands that I can get away with? What’s the lowest signal to noise ratio?  What’s the highest temperature differences that are required?  These are questions that designers need to know in order to lower the cost of the equipment.  This would enable commercial applications to lower their costs of operations.  Is that an appropriate role for NASA?  That kind of research?  Because it almost seems in some sense you could argue that that’s going backwards.

Recommendations to NASA
Research in better thermal infrared capabilities and better calibration.
R: Let me clarify the question.  Thermal infrared (TIR) is not a parallel... It’s very expensive to acquire good thermal measurements, especially if multi-spectral measurements are also required.  The technology to acquire is the most expensive available.  If the applications require very fine measurements with good accuracy, that technology is going to be expensive.  On the other hand, the application requirements can be reduced. It’s possible to take a less expensive approach to thermal measurements aimed at businesses.  If we use the appropriate technology and are willing to take a hit, the question is  “Can you take the hit?”  Who knows the answer to this question?  We’ll have to find out.  And is that an appropriate role for NASA? R: Somebody has to build the expensive system and once you have the data on all of these systems, you can answer the question, what do I truly need?  It’s kind of like hyperspectral systems, there’s no way that you’re really going to use all that data, but you build the system and you tell me which bands I need and I can build a five channel system that I can fly today.  Maybe I’d like to change five bands next week for a different crop but five bands is all I need.  Then it’s a cheap system.  But until somebody builds a system and puts it out there, and lets people use it in practical situations as opposed to esoteric, it’s useless.

P:  I think that people in the business of agriculture or forestry or range management should have an idea of what they want.  We’re writing down questions that are best answered by the little people who know their crops and trees and grasses and that know their problems. R: Hey, soil type and yield are highly correlated. R: So you’re saying that NASA should take on the role of telling agriculturists what they need.  We got into that and that didn’t work. R: Yes, but that was before.  You now have a technology revolution in agriculture that’s come about in the last three years, that has just about changed everything.  R: If I didn’t know that you could get away with a few degrees of temperature difference, I would devise a system that could give me point 1, because I would want to err on the side of being conservative.  That’s going to be the order of magnitude. R: NASA has historically turned its back on agriculture even though agriculture represents the opportunity to change the entire funding situation. R:  You’re wrong.  I spent more of the government’s money on agriculture than the Department of Agriculture did in remote sensing.  How was NASA turning its back? R: NASA spent and wasted huge amounts of money.  R: The solution to NASA’s problems are all solved by GPS and GIS.  The technology needed to solve the problems that NASA’s addressing 10 years ago --these ground sensor systems didn’t exist.  They exist today.   So, you can’t say whether NASA wasted  money or spent it effectively if it was pushing technology without thinking about the entire system.  R: The price of the technology -- NASA may be simply forced for forego developing certain kinds of systems if the conventional wisdom is that what’s required is something terribly expensive.

Potential Clash between Researchers vs. “Small Science”
P:  A word of warning that has to be considered here.  I understand, but it cuts very close to regulations and that will be a real turn-off to a large segment of individuals.  You are going to rapidly run into a situation where researchers have been pushing a particular dogma in the past.  Now you have small science that it’s used to drive regulations and new technology which now gives you an accurate picture and they clash violently.  An example:  If you farm well, whatever well is, and you create large earthworm populations which is often thought to be an indication of soil health and productivity, and you leave lots of residue on the soil, then you measure your residue immediately after cropping, you end up with a certain amount of residue.  During the off season (winter), the earthworms chew up all that residue.  You can go from having a legally good field to having this land that’s grossly out of compliance with current regulations.  The problem is with the way bad science has been used to write regulations in the past.  The easiest way to solve that problem is to spray a chemical that kills the earthworms.  You don’t have any residue degradation during the winter, and everyone is happy.  There is that whole area of science that starts becoming real fuzzy that people have to address or you get a knee jerk reaction and perhaps bad laws.

Soil Moisture Research with SARs
P:  Some problems I’m concerned about here involve what we call precision agriculture.  Technology doesn’t know exactly how much input has been put on a piece of land... R:  So, if I’m going to emphasize soil moisture -- there are any number of sensor systems from visible to microwave.  The ones that are consistently showing the most promise are the SARs. R:  Are you talking about seed soil moisture, surface soil moisture or root soil moisture?  Big difference! R:  All of the above.  With multi-band, multi-frequency SARs you can get a handle on all of these if results to data are an indication. R:  You just defined an argument that would make a project sail through an agriculture review group and see lots of collateral support for NASA funding.

P:  [Muffled] I believe there is a lot of potential out there in models that look at radar of varying penetrations in depth and models along with the precipitation index and the soil types and the curves that you get from different depths as a water balanced accounting model.  I am not emphasizing this because I have no basis except for the results of really old experiments I’ve seen.  I can remember some of the really old experiments with really larger and  broader resolution data looking at soil moisture data.  It could be tried again with finer resolution ... R: Can you fly it on a plane?  Because if you’re trying to answer the problem in agriculture, your soil moisture is changing dramatically from points within meters of each other.

Thermal Satellites
P:  I’d like to go back to thermal.  Thermal aircraft TI<S, which you may know about.  The Mission to Planet Earth is not putting big bucks in it -- definitely not rebuilding it.  We’re going to definitely try to keep it alive until ASTER is built.  MASTER is the ASTER Simulator.  It would be good to have one flying season where we can correlate the calibration and then vary TIMS. R: That’s okay.  I was going to point out that ASTER is a thermal experiment and it’s a really good one apparently. R: MASTER’s been funded. R: That’s great.  You’ve got that first step that you need, as long as you will fly the airplane ... R: There’s no plans to do multi-spectral thermal. R: There’s only four channels in the spacecraft, the simulator’s got to have more than four channels. R: I’m sorry you’re right there is. [Muffled comments]

P:  We’re looking 10 years down the road. R: If you come along and say you want laser for carbon, oxygen, nitrogen, potassium, phosphorus, do you get support?

Need for a “Calibration Vehicle”
P:  I think I would put “Calibration/Atmospheric.”  It better be taken care of simultaneously.  “Calibration of Foreign and Commercial Satellites.” R: [In response to the four questions you brought up at the beginning of the session that NASA had to answer], one possibility is not to build a completely operational LANDSAT 8 system to replace LANDSAT 7.  Focus on providing on some sort of calibration vehicle that would enable NASA to provide a way to focus and compare available data as some sort of yardstick. R: CALSAT! R: If we’re going to have to have a wide swath, we won’t have to worry about ... Just focus primarily ... Take data and prepare it for the system. R: That thing hasn’t been talked about before -- I think it’s a good idea.  NASA has not been thinking in terms of that before.  LANDSAT 8 becomes a question because of Congress.  NASA would build another one or ten more, if they could get the money.  Congress is acting uppity about it for two reasons:  1) they don’t want to spend money, and; 2) they kind of like to think that commercial satellites are going to do everything from now on forever for everyone.  That’s why NASA focuses on so much calibration, accuracy, precision, etc. Which they hope the commercial guys will do.  What they ran into however, with Resource 21, is every precision question they asked Resource 21, what are you going to have in your instruments for sun sensors, etc.  They were doing everything NASA was doing.  The only argument between LANDSAT and Resource 21 was that Resource 21 is going to take the data going up, we’re going to take the data going down.  The data is going to correlate like that.  Big deal, since they’re going to have sample the 10 meter data and the other data and put them together anyway, so that wasn’t a very big issue.  There are some commercial guys -- I just learned today that Lockheed has a sun aperture on their sensor.  Basically with calibration, you can only do a couple of things that everybody doesn’t do.  Everybody has the consistency check inside the satellite.  Using a constant source like the sun is more technology, more “stuff” on the satellite to get that partially through the lens or totally through the lens.  Now we’re down to arguing details.  People in the commercial world are facing up to it apparently.  If that’s true, then NASA has no argument on the quality side.  Nobody in NASA is worried about this problem of calibrating the other instruments, and it seems to me that it’s a good government function.  EOS programs are bombarded with the possibility of using sensors by means by which one sensor could be used to calibrate other systems.  JE: That’s not quite true, the EOS programs and MTPE has a large calibration validation effort that encompasses all EOS sensors. R: Well, they’re going to fly LANDSAT in the same orbit, 20-30 minutes ahead, but that’s sort of a gerrymander.

Argument over NASA’s Role in Remote Sensing
P: When we discussed proper processing of the future LANDSAT data, we’ve run into a lot of… JE:  The difference is you’re talking about commercial applications and products, not science applications and products.  If you could show me an instance where NASA has gotten data satellite acquisition to a farmer in the field within 24 hours, then I’ll say okay. R: I agree with you -- I’m saying that’s an area that needs further research and development. R: But is that NASA’s business anyway?  Because that’s the implementation factor that has to be done by a commercial outfit who may or may not step up to this.  It’s not a problem for NASA to solve, because NASA is never going to do it. R: It is a problem.  It’s an infrastructure issue.  NASA is not going to have these satellites deliver data to the farmer basically. R: Somebody’s going to want to process and analyze the data. R: I think there’s several companies planning to do that. R: When you look at the commercial ones going up they’re worried about it. R: So the problem’s solved.  What’s the government doing in it?  The processing issue will be solved by the proper people.

P:  If I was carving out a role for NASA in agriculture, it is to sit and define the best instruments that can be built and then to publish those results so they can be commercialized relatively inexpensively.  So you cut the barrier price down.  If you want to look at visible band imaging, let’s define the very specific bands that are needed, publish those so that the commercialization can go forward quickly and easily, as opposed to making a very high barrier, because it’s easy to fill an RGB system, but that’s really not what we need. R: Right now with the visible sensors, NASA is not in the forefront.  LANDSAT is because it’s got the calibration, etc.  They’re not in the push broom world very well except they got EO1, which is playing around with that technology for the first time in other nations and other people are flying this and going to fly it before NASA flies it anyway.  So that whole technology (optical technology, analysis of the optical technology) is in the commercial world now.  Several companies are pushing that right now.  They’re going to make it live or die.  NASA doesn’t need to be involved.  NASA needs to be doing things that nobody else will do -- nobody else will develop SAR. R: A lot of people have developed SARs.  Why should NASA do it? R: Not multi-parameter SARs or at a price level that the commercial world will ever be able to use.  There are no commercial SARs up there now.  If you’ve got a $130 million SAR, that’s a giant step forward relative to other government satellites.  And if you’re doing technology transfer, that’s a worthwhile objective, because that gets it down into the private sector price, maybe.  In other things, that are mysteries as to how they can be used or how useful they will be, they all have indicators that they’re good but we don’t have any focus point of data that say just what they’re useful for.  The agriculture people, especially, have never really looked at a lot of this stuff and never had a consistent program to say “This is good.”  There’s been some work in Stennis, and they just broke the Richmond[?] [Soil moisture] is a major area, and the reason the SAR is on this ... Was primarily because of its potential applications.  JE:  I’m not sure the Canadians, the Europeans, the Japanese, or the Russians would agree with you on this SAR issue.  They all have of had SAR systems up there where data was commercially sold.

P: There are two aspects to this question about what NASA should be doing.  One is the capability that previously has not been available to the agricultural community (like multi-spectral).  What capabilities are available that NASA could improve up with respect to performance or cost?  So with hyperspectral we could test different approaches to achieving hyperspectral in remote sensing ... And do some of the research and development and demonstration that are needed.  The question not only is hyperspectral useful, but will it help agriculture enough to make it a greater priority? R: Not in the commercial realm at this time.  R: I would argue the reverse.  RGB provides no useful data to agriculture and without hyperspectral, you might as well do it if you’re going to make remote imaging a viable tool. R: From an operational perspective, hyperspectral is not going to cut it for us.  For a research tool, we can identify where we need to collect more data and then do that ... R: That’s fine.  But I want to ask a very simple question.  I want to separate corn and potatoes.  How many spectral bands do I need to separate those two and which ones are there.  If you can tell me that then we’re further along than I believe we are.  R: I’ll bet you radar will work better than anything. R: Fine!  Just tell me what it is so I can go out and buy an instrument effectively and put it into a field operation.  R: To answer that question, somebody has to do some theoretical calculations and somebody has to have done built and operated a very high resolution, high quality hyperspectral system.  If that data were then made available for everyone commercially to compete for, that would be interesting.  R: You’re all saying the same thing. R: Now there are approaches for how you collect the supporting data that’s not going to make that process very expensive?  R: First it’s required to get this list of all these things that you want to separate. R: There’s 10 million of those issues. R: If NASA is totally incapable of addressing all of that as a mass, you treat it as a block and that’s not satisfactory.  We’ve got to go to this next step where agriculture will tell us where the real issues are.  That’s what the commercial guys are trying to do.  That’s what some groups try to find out. R: Well some want to identify crop types, but others want to know if you have an infestation in the field.  JE:  That’s not our question.  How many hyperspectral technologies exist today or will exist in the next ten years?  You can’t name any.  We’ve got wedge technologies, we’ve got beam splitters, all different kinds of sensor technology -- that’s what we’re going after.  What one doesn’t exist that you want? R: Nothing. R: Right, then ask that question in another group, not here.  Now Jim just said he can think of a few hyperspectral technologies that don’t exist or have not been proven out.  If that’s true, then we’ll add it over there.  If this group gets bogged down with specific questions, we’re going to be here for 20,000 years.

P:  If information on hyperspectral becomes available, which it looks like ultimately it will, then the next issue from practical applications from my perspective is, which is the best type?  How do you trade off different types of sensor design to give you the kind of performance you need?  The companies make those decisions and your stuck with the issues.  You have no input.  NASA is probably the only organization that can objectively define pros and cons of each design within a given set of applications.  Back to consumer reports again. R: Multi-sensor integration, that’s a good one, even the scientists need to do that.

P:  Hyperspectral is a technique of doing the instrument which can be used as a separate band, different bands or anything like that.  You can do all that stuff, the technology is out there.  What you need to do is to find what is really required and then those things have to be identified.  I think we’re getting down to this third or fourth tier instrument design.

Summary of Recommendations for Further Research
JE:  We’ve determined that the limiting factor is not the sensor technology, it’s putting the sensor technology into correct application.  Then we took it a step further.  If there was a weakness in this as it pertains to agriculture, where, given the biases of the people in this room, would we place research needs?  Is at the sensor bands, thermal infrared, SAR, active and passive microwave (for soil moisture), atmospheric calibration or atmospheric correction. R: Aren’t the commercial companies (except for Resource 21) not paying much attention to calibration? R: The guy from Lockheed said they have a sun sensor.  That’s the criteria I use -- if the guy looks at the sun, he’s in the [calibration] club. There’s a really big difference.  JE:  But still this is a major area.  Some people were interested in the area of sensor utilization and some of the laser technology.  They said the other things have been pretty well researched, if you are going to focus on what you needed to do in terms of sensors for agriculture to really beef up a range of things.  There is still a lot to do.  There was a general feeling that microwave is one of the better places to look at soil moisture [research] and that we have never done.  People are doing work with passive and it’s always a very small scale, and from aircraft we may be able to get down to the finer resolutions.  What they’re also talking about is you need soil moisture not only in the surface, but in the seeding area and the root zone.  Either researchers need to make it more available to the broad community or in the open literature and tell us what they’re really doing.  It’s important. R: It was developed to the point where there was control issue ... It never got operational because of vegetation residue, changes in texture, calibrations, so it never advanced to get operational. R: It comes back to calibration and if you look at airplane based data and satellite base systems, producers and agriculture in general, there is no independent way to verify data quality etc. R: Somebody mentioned a CALSAT or putting up a calibrated satellite that would do that and maintain that for the country and for the world.

P: This is a summary of what the other group is saying.  They feel that agriculture, the whole agriculture area is still very premature and still needs a lot of development before substantial  integrated commercial systems can make a substantial investment in them.  JE:  Our group dealt with what is available for use in the next 10 years.  We said Stoney did it with the satellites.  In aircraft, anything you can think of is either available now or will be available in the next 10 years.  A lot of it exists through NASA. R: What they need to hear is the fact that the whole textbook for agriculture has been tossed out in the last two years.  R: We’re really saying two things – you’ve got toys now that need to be exploited and if you do research, they can lead you to those areas that you ought to be focus on.  You’re going to fly a SAR, but you don’t have a lot of work to support the agriculture uses of the SAR -- that’s the one of the big things you neglected over the years.  For agriculture, it has so many dimensions to it -- it just boggles the mind what it’s going to do. R:  As I recall, when NASA put money the thermal and the SAR, it was for geology. R: Yes, it wasn’t for agriculture.  But still, everyone else is using it for agriculture, forestry, etc.; but, we’re not. R: The only commercial venture that’s going to look at agriculture specifically at a targeted market is Resource 21.  Most of the other ones are water based for the next ten years. R:  What I heard from Lockheed today is that Yes, they’re going to go hard on agriculture, but that’s because they can afford it because they’ve got a base set of anchor tenants from which they’re going to make real money.