PRECISION FARMING AND VARIABLE RATE TECHNOLOGY A Resource Guide

Transcription

1 PRECISION FARMING AND VARIABLE RATE TECHNOLOGY First Edition February, 2010 Agricultural Research and Extension Council of Alberta (ARECA) 2010 ARECA

2 Table of Contents List of Tables...iv List of Figures...v Forward... 1 Acknowledgements... 1 Proprietary notice... 1 Conventions... 2 Introduction to Precision Agriculture Technologies... 3 What is Precision Agriculture?... 3 First farmers and small plots... 4 Managing variability... 5 Brief history of Precision Agriculture... 5 GPS as the enabling technology... 5 Early efforts in managing variability... 5 Global Positioning System (GPS)... 6 What is GPS and how it is used in Precision Agriculture?... 6 Types of satellite positioning (GPS, GLONASS, others)... 7 Basic description of how GPS works... 7 Receiver types... 9 Error Sources Differential corrections and accuracy levels Why are corrections needed? Satellite Based Augmentation Systems Beacon RTK Guidance, controlled traffic, other uses of GPS Basic manual guidance Autosteering Controlled Traffic Farming Mapping weeds, rocks, obstacles Automatic section control and mapping Geographic Information Systems (GIS) What is a GIS and how it is used in Precision Agriculture Visualization of layers concept Paper based maps to software packages Layer management Spatial data types, resolution, quality, etc Data types Geo referenced data Garbage in garbage out Data parsing and verification Precision Farming and Variable Rate Technology i

3 Variable Rate Technology (VRT) What is VRT and how it is used in PA Inputs required Yield data Fertility Topography Imagery Electrical Conductivity (EC) Target yield Manual control Map based control Variable Rate Applications Fertilizer Herbicide Fungicide Insecticide Seed rate Desiccation Irrigation Yield Monitoring Managing yield data Accuracy and errors Calibration Grain throughput lag time Weeds and chaff Position errors Remote Sensing Types of remote sensing Satellite imagery Aerial Non contact sensors Google Earth as a data source Spectral Bands: Visible, NDVI, Multi spectral, hyperspectral Visible Infrared, near infrared, false color infrared NDVI Multispectral Hyperspectral Resolution How resolution affects data quality Resolution Classes Ground truthing Economics of Precision Agriculture Guidance Precision Farming and Variable Rate Technology ii

4 Payback through reduced overlap, less waste, less fatigue Operation in poor conditions dark and dusty Round the clock operations to maximize machine productivity VRT...48 Data Requirements Payback through better management More consistent yields Maximizing profitability What if scenarios: what to grow where Inventory management Service providers Application costs Data management Do something today! Websites Selected References Appendix A: Glossary Precision Farming and Variable Rate Technology iii

5 List of Tables Table 1 Differential correction accuracy Table 2 Example equipment for Controlled Traffic Farming Table 3 Sample attributes for a soil sample point Table 4 The Classic Partial Budget Format Table 5 Example Input Data Format Precision Farming and Variable Rate Technology iv

6 List of Figures Figure 1 Precision Farming Cycle... 4 Figure 2 Early farmer harvesting his crop... 4 Figure 3 GPS Satellite... 6 Figure 4 GPS satellite constellation... 7 Figure 5 Trilateration from 3 GPS satellites... 8 Figure 6 Example of satellite visibility and DOP from planning software... 9 Figure 7 GPS signal error sources Figure 8 Dilution of Precison Figure 9 Repeatable accuracy of GPS with various corrections Figure 10 Conceptual overview of WAAS Figure 11 RTK Base Station Figure 12 Example of a lightbar guidance system Figure 13 Example of an onscreen guidance system Figure 14 Steering wheel mount autosteer Figure 15 Tilt compensation on a tractor Figure 16 Roll, Pitch and Yaw of Tractor Figure 17 Controlled Traffic Farming using a 3m track width Figure 18 Controlled Traffic Farming with different track widths Figure 19 Example of picking rocks with GPS guidance Figure 20 Automatic section control Figure 21 Layers form the information in a GIS Figure 22 Line example represented by several points connected together Figure 23 Polygon example representing a field boundary Figure 24 Example of soil types overlaid on satellite imagery from Google Earth Figure 25 Example of a geo referenced field border Figure 26 Example yield map Figure 27 Example of zone soil sampling Figure 28 Example of grid soil sampling Figure 29 Example 3D elevation map Figure 30 Example of raw and analyzed satellite imagery Figure 31 Example of NDVI image used to create VRT zones Figure 32 Example of electrical conductivity soil sensor Figure 33 Example of an electromagnetic conductivity meter Figure 34 Example of map based variable rate control in sprayer Figure 35 Example fertilizer prescription map Figure 36 Example of variable rate irrigation Figure 37 Example of yield data before and after cleaning data Figure 38 Remote sensing unit on a fixed wing aircraft Figure 39 Examples of non contact crop sensors Figure 40 Visible spectrum Figure 41 Infrared spectrum Precision Farming and Variable Rate Technology v

7 Figure 42 Example of an NDVI field image Figure 43 Example of Hyperspectral Remote Sensing Figure 44 Example of 30 m, 10 m and 2.4 m resolution satellite imagery Figure 45 Example of a profit map Precision Farming and Variable Rate Technology vi

8 Forward The purpose of this manual is to serve as reference material for those attending precision farming training workshops hosted by ARECA and its member associations in Alberta. Technology changes quickly and the materials discussed are believed to be current at the time of production. No representation or warranty is made as to the accuracy or suitability for your particular circumstances. Readers are advised to consult with qualified practitioners before applying the methods described in this manual to their operation. Acknowledgements The Agricultural Research and Extension Council of Alberta would like to thank the following contributors for making this project successful: Alberta Agriculture and Rural Development Growing Forward Program for financial support of this program Dale Chrapko (Alberta Agriculture and Rural Development) for financial support Doug Mackay, P.Eng. for development of this manual Jay Bruggencate and Colin Bergstrom (Farmers Edge) for imagery contributions Ted Darling for economic analysis contributions Warren Bills (GeoFarm) for imagery contributions Proprietary notice Copyright 2010 Agricultural Research and Extension Council of Alberta. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means without prior permission of the copyright owner. Other trademarks and trade names used in this manual are the property of their respective holders. Agricultural Research and Extension Council of Alberta #211, 2 Athabascan Avenue Sherwood Park, Alberta T8A 4E3 Tel: Fax: Website: faechner@areca.ab.ca First Edition February, 2010 Precision Farming and Variable Rate Technology 1

9 Conventions The following conventions are used in this manual: Italics Sidebar tips The term in italics is defined in the glossary at the end of this manual TIPS: Helpful advice and things to watch out for. Precision Farming and Variable Rate Technology 2

10 Introduction to Precision Agriculture Technologies What is Precision Agriculture? Precision agriculture attempts to manage variability within fields. There are several reasons that precision farming has come about as a management method in the past decade: High cost of crop inputs including seed, fertilizer, pesticides and fuel Environmental concerns about fertilizers and pesticides near sensitive areas, runoff and de nitrification The technology has become available and economically feasible Precision agriculture goes by many names but they all refer to managing variability: Precision farming GPS farming Prescription farming Farming by satellite Spatially variable agriculture Farming by the foot Site specific management Variable rate application Just as the seasons have a cycle, so does precision agriculture. Throughout the season, data is collected from various sources, analyzed, and used for decision making. There are many types of information that can be collected and used in a precision agriculture system including yield, topography, prescriptions, imagery, electrical conductivity, soil types, and as applied maps. Precision agriculture is really a systematic approach to managing fields. It involves multiple technologies and multiple disciplines. It should be first and foremost considered a tool to make better management decisions. It is not a silver bullet that will make any farm profitable but it will help with making more informed decisions that can lead to greater profitability. Precision Farming and Variable Rate Technology 3

11 Figure 1 Precision Farming Cycle Source: Virginia Cooperative Extension First farmers and small plots Early farmers managed their land intimately. They walked the fields to cultivate, sow, remove weeds, water, and harvest their crops. They intuitively knew that some areas of the land yielded better than others. Much of this information was retained in their memory or recorded in a notebook. This information may have been passed down from one generation to the next verbally or in written form. Figure 2 Early farmer harvesting his crop Precision Farming and Variable Rate Technology 4

12 Managing variability As modern agriculture led to larger and larger machinery and the number of acres grown by individual farms increased, the intimate knowledge of the land was lost, or at least much more difficult to manage. For many years equipment operators have manually adjusted the spray rate when driving through a heavily infested area but this can be fatiguing and not 100% accurate. Modern technology allows automation of these tasks. Brief history of Precision Agriculture GPS as the enabling technology Without having a reliable method of locating equipment and items in a field, it is difficult to manage in field variability. A crude method might be to stake out the field to show areas that require different treatment, but this is not practical on large fields. A reliable positioning method is needed to accurately locate field features to make precision agriculture work. Some local positioning systems were developed but not successfully commercialized. The advent of GPS allowed for low cost, reliable positioning of equipment in the field. Data from other sensors could be tied to a specific point in the field with precision. Early efforts in managing variability Various methods of precision farming have been tried over the years, with varying degrees of success. An operator that is very familiar with a field and able to detect differences in soil type, topography, yield potential, etc., may be able to adjust his rates on the go while travelling up and down the field. Some of the problems with this method is that is rapidly fatiguing for the operator to be constantly watching and adjusting rates. Also, there is no feedback to record what was applied where to verify if the treatment was cost effective and profitable. Precision Farming and Variable Rate Technology 5

13 Global Positioning System (GPS) The Global Positioning System (GPS) was launched by the United States government initially as a way to locate military applications. It has grown into a commonplace utility, being used in everything from cell phones to cars to landing aircraft and guiding ships. Figure 3 GPS Satellite Source: USAF GPS factsheet What is GPS and how it is used in Precision Agriculture? We said earlier that GPS is the enabling technology of precision agriculture. It is the positioning system of choice since it is reliable, free to use with the correct receiver, and relatively accurate. Since the early 1990 s, producers have been using GPS in field applications. Uses of GPS in precision agriculture: Yield mapping Autosteering Variable rate control Automatic section control Field mapping Drainage Guidance Data collection Asset tracking Crop scouting Irrigation Topographic mapping Tracking livestock Electrical conductivity mapping Aerial spraying Soil sampling Rock picking New uses are being developed Precision Farming and Variable Rate Technology 6

14 Types of satellite positioning (GPS, GLONASS, others) There a two major satellite positioning systems circling the earth. The main system used in North America is the US military Navstar GPS system. There is also the Russiandeveloped GLONASS system, which can be used by some receivers in North America. Systems under development include the European Galileo system, Chinese Beidou and Compass, and India s Indian Regional Navigational Satellite System (IRNSS). These are mainly intended to be local systems for specific areas of the world to limit dependence on the US or Russian systems. The general term for all satellite positioning systems is Global Navigation Satellite System (GNSS). Basic description of how GPS works The Global Positioning System consists of three parts: Space Segment: Control Segment: User Segment: 24 operating satellites in orbits about 20,200 km above the earth that pass by twice per day (12 hour orbits). Monitor and control stations on the ground that maintain the satellites in their proper orbits and adjust the satellite clocks to maintain the health of the GPS. The user s GPS receiver on the ground. This can be in the cab of the tractor, an airplane or a handheld device. Figure 4 GPS satellite constellation Source: Garmin GPS works by transmitting radio frequencies from the orbiting satellites. These frequencies contain information on the satellite s health, location and a timing signal that is received by the user s GPS on earth. The constellation is arranged so there at least 5 to 8 satellites visible anywhere on the earth at any given time. The GPS receiver Precision Farming and Variable Rate Technology 7

15 on the ground calculates its location by measuring the time it takes the signal from the satellite to reach it. By calculating the distance to several satellites, the GPS receiver can determine its position. With 3 satellite signals, the GPS receiver locates itself at the point where the circles of the distances from the 3 satellites intersect. This process is called trilateration because the GPS receiver is measuring its distance (lateration meaning distance to a point on a sphere) from at least 3 satellites (tri meaning three). A minimum of four GPS satellites are needed to receive a three dimensional (3D) position (latitude, longitude, altitude). The more satellites that can be received at one time, the less uncertainty there is in the position. Figure 5 Trilateration from 3 GPS satellites There is free software available to visualize how many satellites are available at any time and predict HDOP at your location. Since the GPS system is now fully operational, there are usually enough satellites available overhead but if there are obstacles in the way, this can be reduced. Most receivers also set an elevation cutoff angle at which satellites are not used if they are a certain elevation off the horizon, as their accuracy is decreased since the signal travels through more of the atmosphere. Typical elevation cutoff angles are 5 10 degrees. Where can I download planning software? planningsoftware.shtml Current almanac data: gpsdataresources.shtml Precision Farming and Variable Rate Technology 8

16 Figure 6 Example of satellite visibility and DOP from planning software Source: Trimble Navigation The whole process of how GPS works is quite complex and the reader is welcome to explore some of the websites at the end of this guide for more information. Receiver types GPS receivers come in different accuracy classes and utilize different correction sources. The accuracies range from recreational class non corrected receivers at 15 m to survey grade professional equipment that can attain millimeter accuracy. Of course, the cost of these receivers varies widely with performance and features. Most modern receivers have at least 12 channels to receive signals from 12 GPS satellites at one time. Some have more, as many as 50 channels in anticipation of utilizing GPS, GLONASS, and SBAS signals at the same time. There are single and dual frequency receivers. Single frequency receivers are generally less accurate than dual, but they are also less expensive. Dual frequency is used in the highest accuracy RTK applications of 1 accuracy or better. Dual frequency RTK receivers have faster initialization times than single frequency. GPS receivers can have various differential correction methods built into the receiver. L Band corrections can use the same channels as the GPS receiver. Beacon and RTK require a different frequency and may have the radio receiver built into the GPS receiver box or a separate radio receiver. Precision Farming and Variable Rate Technology 9

17 There are new GPS frequencies coming in the next few years called L2 and L5 that will enhance the accuracy and strength of non corrected GPS signals. This is meant to enhance GPS for safety of life applications such as locating cell phones in urban areas. Error Sources Ionosphere There are several sources of error with GPS signals. The radio signal travels from satellites some 20,000 km away basically at the speed of light until reaching earth. Once the signal reaches earth, it must pass through the ionosphere which is an area of charged particles about km above the earth. The signal is delayed as it travels through this region and the delay is frequency dependant. Dual frequency receivers can virtually eliminate this error by measuring the difference in the delay. What are the sources of GPS errors? Orbit error ±2.5 m Satellite Clock ±2 m Ionosphere ±5 m Troposphere ±0.5 m Multipath ±1 m Receiver noise ±0.3m Troposphere The next layer is the troposphere, which is about 20 km above earth. The signal delay through the clouds and moisture in the troposphere is not frequency dependant but can be modeled to compensate for it. Obstructions Signals can be obstructed by buildings, trees, or metal objects. This is why it is important to mount the GPS antenna with a clear view of the sky on the equipment so that there is a better chance of receiving all the satellites that are available. Multipath GPS signals can bounce off objects near the antenna, causing a delay in receiving the signal. Many modern antennas have built in multipath rejection. For high accuracy antennas a choke ring device can be added that helps to reject multipath. Satellite clock The GPS satellites each have atomic clocks onboard for precise timing. The GPS receiver on the ground has a lower accuracy quartz clock. There are processing methods that increase the timing accuracy of the receiver s clock but still does not have the accuracy of an atomic clock. Considering the signal from the GPS travels at the speed of light, even a nanosecond of timing error amounts to a measurable error. Precision Farming and Variable Rate Technology 10

18 Figure 7 GPS signal error sources Dilution of Precision (DOP) An important factor in calculating the accuracy of GPS is called Dilution of Precision (DOP). It is a geometric factor that serves as a multiplier of the general accuracy of the receiver. There are several types of DOP, including horizontal (HDOP), vertical (VDOP), time (TDOP), 3D position (PDOP) and geometric (GDOP). When the satellites are widely spaced around the user, the DOP value will be low. When the satellites are clustered together the DOP value will be higher. The DOP value is multiplied by the base accuracy to get the current accuracy of the receiver. HDOP values under two are considered acceptable. Figure 8 Dilution of Precison Precision Farming and Variable Rate Technology 11

19 Differential corrections and accuracy levels GPS signals alone are typically ± 15 meters. While this is fine for locating a ship in the ocean or a hiker finding their way, it does not provide the accuracy needed to guide equipment in the field or collect topographic data. Why are corrections needed? Without corrections, GPS would not be useable for many of the applications in precision agriculture. With 1 3 m pass to pass accuracy and 15 m repeatable accuracy, uncorrected GPS might get the user to the correct field, but not much more. Guiding equipment to sub inch accuracy only works with sophisticated correction signals being applied. Table 1 Differential correction accuracy Source: Trimble Navigation, Deere & Co, USCG, OmniSTAR Correction Type Pass to Pass Accuracy Repeatability Non corrected ± 1 3 m 15 m WAAS ± Beacon ± m (Decreases with distance from base station) OmniSTAR VBS ± 6 8 < 1m OmniSTAR HP ± OmniSTAR XP ± StarFire SF1 ± StarFire SF2 ± 4 10 StarFire RTK ± 1 1 RTK ± 1 20 cm 1 20 cm (Decreases with distance from base station) Figure 9 Repeatable accuracy of GPS with various corrections Precision Farming and Variable Rate Technology 12

20 Satellite Based Augmentation Systems Satellite Based Augmentation Systems (SBAS) provide corrections to GPS receivers via an L Band satellite signal separate from the GPS signal. It is used to improve the accuracy of the user s location to sub meter or better. WAAS The Wide Area Augmentation System (WAAS) was established by the United States Federal Aviation Administration (FAA) for aircraft approach positioning. There is no charge to use the WAAS correction signal and many receivers have built in WAAS capability. WAAS uses geo stationary satellites over the equator on the east and west sides of North America to provide coverage of most of the continent. There is similar system in Europe called European Geostationary Navigation Overlay Service (EGNOS) operated by the European Space Agency. Figure 10 Conceptual overview of WAAS Source: USCG Navigation Center OmniSTAR OmniSTAR is a commercial provider of SBAS signals. They offer three levels of accuracy from 2 to 8 by subscription. They broadcast on an L Band frequency and require an OmniSTAR capable receiver. Precision Farming and Variable Rate Technology 13

21 StarFire John Deere operates their satellite correction signal that works with their GreenStar system. They offer three accuracy levels, two via satellite correction and one via an RTK base station (source: Deere & Co.). John Deere dealers have also set up their own RTK network, available by subscription, which provides the user with corrections without having to set up their own base station. Beacon The US Coast Guard and Canadian Coast Guard set up DGPS reference stations to assist in waterway navigation around coastlines and large lakes. While this system is more widely available in the United States, coverage in Canada is limited. The United States Department of Homeland Security has expanded coverage to include the entire inland United States, but Canadian coverage is still limited. Beacon corrections are free to receive with the proper GPS receiver. The accuracy of beacon corrections decreases with the distance from the reference station. While the corrections may be received for up to 250 to 350 km from the station, accuracy decreases the further away the receiver is from the reference station. Testing is being done to increase the accuracy of the beacon corrections in the United States to 10 cm. (Source: US DOT) RTK In order to get a repeatable high accuracy position, a method of differential corrections known as Real time Kinematic (RTK) was developed, mainly for the survey industry. The application of RTK to agriculture has meant very high accuracy positioning down to the subinch level. The accuracy of RTK corrections depends on the accuracy level of GPS receiver on the vehicle and the distance from the base GPS receiver. The further away from the base station, the lower the accuracy will be. For best accuracy, the RTK base station should be less than 10 km from the vehicle in the field. Accuracy begins to degrade after this distance but might still be sufficient. What accuracy do you need? Generally, cost goes up with accuracy. Some applications need to be more accurate while others can do with less. Locating the correct field: 1 3m Locating rocks and field features: 1 3m VRT application: submeter Automatic section shutoff: sub foot Manual guidance: sub foot Autosteering: < 6 to sub inch Inter row seeding: <2 Drainage: <2 to sub inch RTK Considerations Cost of setting up your own base station vs. purchasing an annual subscription How many units will be used at the same time? Is there a benefit to you having a high speed internet connection in the cab? Line of sight issues are there obstructions, lack of cellular coverage in your fields? Precision Farming and Variable Rate Technology 14

22 RTK requires a real time radio link from a base station set up over a known location to the moving receiver in the vehicle. There are different methods for receiving the signals from the base station. Local base station A base station can be set up near the field over a known location to provide corrections to equipment in the field over a radio link. The base could be located at the farm on a high point of a building or grain bin to give maximum distance. The radio link requires line of sight to the receiver so the higher the transmitting antenna, the less obstacles will block the signal. It is also possible to set up a radio repeater in the field if reception is an issue. Figure 11 RTK Base Station Source: Trimble Navigation RTK via cellular data Another method of receiving RTK corrections is over a cellular network. The range of receiving these signals is generally further than a single base station with a 900 MHz radio link. It is not line of sight dependent and works wherever a cell phone signal can be received. It will not work in areas of poor cellular coverage. Some providers of RTK over cellular also feature a high speed internet connection that can be used for transferring data, remote job setup, internet access, asset tracking, , etc. This can be a powerful tool and levels the playing field for farmers and urban dwellers by providing wireless high speed internet to rural residents. Precision Farming and Variable Rate Technology 15

23 The range of cellular based RTK can be greater than a single reference station because it is not limited by radio reception but by cellular coverage. Often the RTK base stations are part of a network that models the corrections over a larger area than a single base station, allowing for greater accuracy than with a single local base station. The accuracy may be greater by using a network of base stations as the atmospheric corrections can be modeled more accurately in three dimensions by the network rather than just a single dimension with a single base station. A government initiative mostly in the United States has set up a network of reference stations known as Continuously Operating Reference Stations (CORS). Correction data is available over the internet and requires a cellular data connection in the cab to receive the corrections. Some CORS stations are free to receive data while some charge a subscription fee. There are also the cellular data charges to receive the correction data. Guidance, controlled traffic, other uses of GPS With special processing techniques, GPS can be used to accurately guide equipment in the field. GPS guidance can minimize overlap and misses in the field during application, making more efficient use of crop inputs. Fatigue is reduced since the operator can be confident that they are getting complete coverage. Pass to pass vs. repeatable accuracy It is important to understand the difference between pass to pass and repeatable accuracy as they are quite different, depending on the correction source. Pass to pass accuracy means the accuracy the user can expect over a short period of time, from one pass down the field to the next. Non RTK correction sources are often in the 2 8 range pass to pass. RTK can be from 1 20 cm pass to pass, depending on the receiver. Repeatable accuracy refers to the accuracy the user can expect if they return to the same point in the field the next day or a year from now. Repeatable accuracy of non RTK correction sources varies from 4 to 3. RTK can be from 1 20 cm repeatable. It is important to make the distinction between pass to pass and repeatable accuracy when shopping for a GPS system. If the user needs to return to the same spot year after year with high accuracy, then they need to consider an RTK system. For example, for inter row seeding, high pass to pass accuracy and high repeatability are needed. In a 2009 Purdue University/CropLife magazine survey in the United States, 92% of respondents were using some sort of GPS guidance for custom application and 56% were using auto control/autosteering (Whipker and Akridge 2009). The survey was of crop input dealers and custom applicators in the United States. There has been a rapid uptake of GPS guidance by producers in Canada as well. Precision Farming and Variable Rate Technology 16

24 Basic manual guidance There are two main forms of manual guidance that give the operator cues to steer the vehicle to keep on a specified track. Lightbar The simplest of manual GPS guidance methods is a row of lights in the operator s field of view that indicate which way to steer to stay on course. The operator tries to keep the light in the middle of the lightbar. Some lightbars offer text displays and recording of field features that can be navigated back to at a later time. This is useful for returning to the last point that was sprayed when leaving the field to refill the sprayer, as an example. Figure 12 Example of a lightbar guidance system Source: TeeJet Technologies On screen guidance Graphical guidance systems provide more information to the operator on what is around him in the field. This is known as situational awareness and comes from technology used in fighter jets and space shuttles to show the user what is around them, even if they can t see it. Figure 13 Example of an onscreen guidance system Source: Raven Industries Precision Farming and Variable Rate Technology 17

25 Autosteering Automatic steering of the vehicle in the field can reduce fatigue and allow longer operating hours while keeping the vehicle on a precise track. There are two primary type of autosteering systems, steering wheel mount and hydraulic. Steering wheel mount An electric motor can be attached to the steering wheel of the vehicle that is controlled by the GPS and electronic controller in the cab. Some attach directly to the steering wheel shaft while others have a roller that turns the outside of the steering wheel. The steering wheel mount autosteer provides medium accuracy. It is more accurate and less fatiguing than manual guidance but not as accurate as hydraulic autosteering, although their accuracy is improving with time. Figure 14 Steering wheel mount autosteer Source: AgLeader Hydraulic Hydraulic autosteering ties directly to the oil lines of the vehicle s steering system. It is generally more accurate than a steering wheel mount as there is no slippage on the wheel and more precise adjustments can be made. The wheels are being controlled directly through the hydraulics rather than through the steering wheel into the steering box which then controls the hydraulics. There are generally two types of steering valves used, proportional and on off or bangbang. The proportional valve opens in response to an electrical signal that opens the valve based on the amount of current or voltage fed to it. For 100% of the current or voltage fed to the valve, the oil flow will be 100%. If the signal is less than 100%, the valve will open proportionally to that amount. For example, if the signal is 50% the valve will open half way. The control of the valve is handled by the electronics in the autosteering module and requires calibration for each machine it is installed on. Precision Farming and Variable Rate Technology 18

26 The on off type valve simply opens to 100% whenever an electrical signal is sent to it and is closed when there is no signal. This type of valve can be pulsed to change the flow rate. A faster pulse will create smoother steering inputs than a slower one. Many of the current autosteering controllers can interface directly with the vehicle s electronics via CANBus, ISOBus or ISO 11783/J1939 standards to allow direct connection without having to add a hydraulic block to the vehicle. If the vehicle already has the capability to be steered electronically, adding a controller and GPS can easily make it autosteer capable. Implement steering On hillsides the implement tends to drift down the slope, even though the towing vehicle maintains a straight path. Methods Pros Easily installed Easily transferred between machines Works with all types of equipment Hydraulic Pros Comparing steering wheel mount to hydraulic autosteering Steering Wheel Mount Easily installed if CANBus connection available Available as factory option on some vehicles Can provide sub inch accuracy with RTK GPS Cons Can add to cab clutter May not be as accurate as hydraulic type Can be noisy Cons More complicated installation if hydraulic block required Requires a separate hydraulic block for each machine Not available for all equipment have been developed to control the path of the implement more precisely. This can be done by placing a second GPS receiver on the implement to precisely locate it or by sensing the position of the hitch. The guidance system compensates for the downhill offset by either steering the tractor uphill to account for the draft or steering the hitch to move the implement back into the correct location. Implement steering can also be used for between row seeding and cultivation. By seeding between the rows, less horsepower is required, so a larger seeder can be pulled by a smaller tractor, the seed germinates more quickly, and there may be less disease pressure and erosion. (Source: Australian Government Grains Research and Development Corporation) Terrain compensation The GPS antenna in usually mounted on the roof of the cab of the machine it is installed on to get the best signal reception. When the vehicle drives on a slope, the GPS location will shift down the slope relative to where the vehicle is actually located. This can be compensated for with a terrain compensator that reads the angle of the vehicle from level and adjusts the GPS location accordingly. There are several types of terrain compensators available, ranging from single axis level detectors that measure roll to 6 axis gyroscopes which adjust the position for side to side (roll), front to back (pitch), and skew (yaw). Precision Farming and Variable Rate Technology 19

27 Figure 15 Tilt compensation on a tractor Source: Terradox Corporation Figure 16 Roll, Pitch and Yaw of Tractor Source: Trimble Navigation Controlled Traffic Farming The concept of keeping machinery wheels in the same tracks year after year is not new. Tramlines have been around for a long time and involve blocking off a shank on the seeder directly behind the tractor wheels to give a visual reference. The new development is using high precision GPS to allow farmers to match their equipment size to the same wheel spacing, typically 3m (9.8 ft) and consistently travel the same wheel paths on each operation. Using a 3 m spacing on tractors, sprayers, combines, etc. allows the equipment to still travel on roads. Australian research suggests that compaction from tractor wheels caused 15 30% yield loss at a cost of $300 $450 AUS in 2007 (White 2007). Precision Farming and Variable Rate Technology 20

28 An example of a farm using controlled traffic might have the following equipment: Table 2 Example equipment for Controlled Traffic Farming Machine Size Multiple of base width Seeder 30 or 60 ft 1 2 Sprayer 90 or 120 ft 3 4 Swather 30 ft 1 Combine 30 ft 1 Figure 17 Controlled Traffic Farming using a 3m track width Source: CTF Europe All the machines would have their wheel tracks set to the same width to minimize compaction. Often the seed rows in the wheel tracks are blocked to save seed as these rows will be trampled anyway. These road beds provide a visual reference for following the rows and high accuracy GPS guidance is typically used to keep straight on the rows. Precision Farming and Variable Rate Technology 21

29 Figure 18 Controlled Traffic Farming with different track widths Source: CTF Europe The above example uses a 12 m (39.4 ) spacing for the seeder and combine and a 36 m (118 ) sprayer. The wheel tracks vary slightly with the combine having a 2.83 m (9.3 ) wheel track and the tractor and sprayer having a 2.2 m (7.2 ) wheel spacing. In this case, tramlines are only used on every third row to provide guidance for the sprayer. Equipment modifications are often needed to convert to a controlled traffic system. Tractor axles may need to be widened and the front axle may require heavier components to account for the wider span. The combine unloading auger may require extension to allow it to reach the grain cart while allowing the tractor to travel on the defined road bed spacing. What about update rates? Update rate refers to the number of times per second a new position is provided by the GPS receiver. For applications like yield monitoring, 1 time per second or 1 Hertz (Hz) is likely sufficient. For guidance applications, an update rate of 5 or 10 Hz is needed. Some high end auto guidance applications may use 20 Hz or higher to ensure smooth tracking. Higher speed applications such as selfpropelled and aerial sprayers require higher update rates than lower speed seeding or harvesting. Precision Farming and Variable Rate Technology 22

30 Mapping weeds, rocks, obstacles While driving the field, the operator can mark the location of field features with GPS. This can be used to mark rocks, weeds, obstacles, irrigation pivot towers and many other features. This allows the operator to return to these locations at a later time. In the case of rocks, he could send another person to the field to the exact location of the particular rock to pick it. Figure 19 Example of picking rocks with GPS guidance Source: Terradox Corporation Automatic section control and mapping Automatic control of sprayer, seeder, and applicator sections down to individual rows or nozzles is now possible. These devices can automatically turn individual nozzles or sections off when they come to the end of the row or encounter a previously applied area or a no spray area. This minimizes the amount of overlap and can increase profits by saving on seed, spray, and fertilizer costs. It also reduces the environmental impacts of agriculture by turning off spray sections in areas that are not to be sprayed. Figure 20 Automatic section control Source: Raven Industries Precision Farming and Variable Rate Technology 23

31 Geographic Information Systems (GIS) Geographic Information Systems (GIS) are used in many industries around the world to collect and manage geographically referenced (geo referenced) data. They serve a useful purpose in precision agriculture to manage and visualize the tremendous amounts of data involved. What is a GIS and how it is used in Precision Agriculture The most basic concept of a GIS is layers of information that can be compared to see trends in the data. A simple visualization is maps of a field printed on transparencies that show the field boundary, soil types, fertility, topography, and yield. If you overlay these layers on one another you may be able to see trends. For example, there may be correlation between yield and topography or yield and fertility. Visualization of layers concept A GIS allows the computer to do the visualization by comparing layers and reporting the correlations and differences between them. Figure 21 Layers form the information in a GIS Source: SST Software Paper based maps to software packages A GIS can be as simple as a book of printed maps showing things like aerial photography and soil types. These printed maps can be manually interpreted with simple overlays of transparencies drawn on with markers. Precision Farming and Variable Rate Technology 24

32 A GIS can also be complicated software programs with complex algorithms that automatically correlate data layers and do much of the thinking for the user to produce output such as a variable rate prescription. Layer management There can be several layers of information used in a GIS. While the computer can handle massive amounts of data, it isn t that smart at interpreting it. A human with knowledge of what they are looking for is needed to wade through the data and find answers to their questions. Layers can be overlaid on each other and compared to look for correlations between them. By comparing layers the user may get answers to some of their questions such as: why was the yield low in a certain area? If they compare to the topographic layer they may see that it was a low depression that drowned out. There may be correlation between fertility and electrical conductivity readings or between elevation and yield. Spatial data types, resolution, quality, etc Data types Points A point is any single location marked with a coordinate. Examples of points could be rocks, trees, center pivot towers, grain bins or other small features that can be defined by a single coordinate. Most GIS data is collected on a point by point basis. Yield or topographic data represents the reading at the specific coordinate it is taken from. Lines Lines are several points joined together to connect the dots. Curved lines can be created from multiple straight lines. A coarse curved line would only have a few points defining it. A fine curved line would require many points to smooth out the line. Attributes can be tied to lines such as width, height, type, etc. Examples of lines include irrigation pipes, roads, ditches, and fences. Guidance trajectories and vehicle paths are also represented as lines. Figure 22 Line example represented by several points connected together Precision Farming and Variable Rate Technology 25

33 Polygons A polygon is a closed area. It is basically a line with the last point joining to the first point. Polygons are used for field boundaries, prescription zones, and soil zones. They could also represent buildings, sloughs, grain storage, etc. They have attributes tied to them such as area, rate to apply, yield, etc. Figure 23 Polygon example representing a field boundary Attributes An attribute is the information that is tied to a specific data point. Attributes are what makes GIS a powerful tool for data analysis and management. Attributes can be any information that the user wants to collect in his database. For example, a soil sample location might have the attributes of N, P, K, ph, organic matter, CEC, micronutrients, etc. Each sample point in the field will have these attributes collected. Table 3 Sample attributes for a soil sample point Data Point Attribute Value 1 N P 7 K 19 ph 5.5 CEC 15 Shapefiles A common format for GIS data is called a shapefile, which was developed by ESRI for use in their GIS programs. It is commonly used in GIS programs although newer formats are available. There are three files required to make a complete shapefile set. The main file has the extension.shp and contains the feature geometry. The second file is the index file with a.shx extension and is used to index the feature to allow faster searching Precision Farming and Variable Rate Technology 26

34 within the shapefile. The third file is a dbase III format.dbf file and contains attribute information. These three files are all required to have a complete shapefile set. There are also optional files that can contain the projection and other information. KML for Google Earth Map data can be displayed in Google Earth when using a KML file. Many GIS packages have the option to save data in KML format. Google Earth has many GIS type features allowing you to overlay images over the satellite image, create points, lines and polygons, and measure areas and distances. As a free basic tool for precision agriculture, it has practical use. How do I get Google Earth? Google Earth is available as a free download from earth.google.com Figure 24 Example of soil types overlaid on satellite imagery from Google Earth Source: Google Earth/CanSIS Geo referenced data Looking at a several paper maps at one time, for example, an aerial photo, a soil classification map and a yield map is a lot of information to mentally process. If these maps are at different scales, which they are most likely are, a GIS system would have difficulty matching them up. By knowing the location of each point on the map, it is easy for the GIS to correlate individual points from each layer of information. Geo referencing refers to tying a coordinate to each point of the data file. This allows easier manipulation of the data and allows scaling of layers to match up with different layers of data. Precision Farming and Variable Rate Technology 27

35 Geo referencing can be done in several ways. An aerial photo can be referenced by taking a GPS reading at identifiable features on the photo such as the corners of the field, power poles, building locations, etc. By tying these coordinates to the features on the photo, other points can now be interpolated in the GIS to show the coordinates of any point. Many forms of data are automatically geo referenced when they are collected, such as yield monitor data, as applied maps and topography data. Satellite imagery is also georeferenced as is most modern aerial imaging. Aerial imagery taken in the past may not be geo referenced but can be brought into the GIS by matching it up to previously georeferenced layers or acquiring the coordinates of features visible on the image. Figure 25 Example of a geo referenced field border Source: Farmers Edge Garbage in garbage out The quality of the data collected determines the quality of decisions that can be made. For example, if the yield monitor was not functioning correctly and giving erroneous data, decisions to create a VRT prescription from the yield data will not be reliable. It is important to verify data accuracy through field scouting and by checking and calibrating equipment regularly. Data parsing and verification There will occasionally be extra data in the GIS file that needs to be removed to have reliable data. This could be data from outside the field, GPS blunders that throw the position out by potentially several kilometers, or incorrect readings. Some of this is done automatically by the GIS software but there could be undetected errors in the data that should be verified. This can be time consuming and is one of the ways that a skilled consultant can benefit your operation. Their skill and knowledge can help prepare better quality data than an unskilled novice could do. Precision Farming and Variable Rate Technology 28

36 Variable Rate Technology (VRT) What is VRT and how it is used in PA Variable Rate Technology, VRT for short, combines GIS, GPS, and electronic controllers in the cab to change the rate of any product being applied in the field. In general terms, VRT is accomplished by developing a prescription map, transferring it to controller in the cab of the vehicle, driving the field with the controller changing the application rate based on the prescription map, and recording how much was applied where. VRT can also be done on the fly with sensors that measure what is needed by the crop and adjust the rate accordingly in real time. Inputs required Yield data It is recommended to collect a few years of yield data to get a sense of the trends within the field. As water is typically the most limiting factor to crop production, the yield may vary widely in a dry year versus a wet year. Using only one year of yield data may not provide the necessary information to make informed decisions. It is recommended to view the trends year over year. Figure 26 Example yield map Source: Farmers Edge Precision Farming and Variable Rate Technology 29

37 Fertility An accurate assessment of the nutrients available in the field is important for varying fertilizer rates. Soil fertility can be determined by soil sampling. Where to sample and how many samples to take requires some planning. Two different approaches to soil sampling are grid sampling and benchmark or landscape sampling. Grid sampling refers to taking soil samples at a regular spacing interval, such as every 2.5 acres or 5 acres. Benchmark or landscape sampling is done by determining beforehand where to sample based on imagery or a zone map. Samples are taken from representative areas of the field, for example of the same soil type or slope. Several cores may be mixed to make a composite sample for each zone. Recording the GPS location of the soil sample allows a map to be produced of soil fertility. A branch of statistics called geostatistics can be applied to interpolate between the points. It is possible to return to the same sample points from one year to another if they are geo referenced. Figure 27 Example of zone soil sampling Source: Farmers Edge Precision Farming and Variable Rate Technology 30

38 Topography Figure 28 Example of grid soil sampling Source: GeoFarm Field topography can have a significant impact on yields. Hill tops generally yield differently than low lying depressions. Using topography as an input for determining zone maps can be beneficial. Collecting high accuracy topography data may require a special trip across the field with RTK differential GPS to get 1 or better accuracy. Using a WAAS corrected receiver will not give accurate enough elevations for most uses. For drainage purposes, accurate elevations are needed to ensure the water flows in the correct direction. Figure 29 Example 3D elevation map Source: Farmers Edge Precision Farming and Variable Rate Technology 31

39 Imagery Imagery acquired from satellites or aerial means can be used to determine where inputs should be applied. Imagery refers to film based aerial photos, satellite digital images or any other overhead view of a field, generally taken looking directly down at the field. There are several types of imagery that are described later in the Remote Sensing section. Figure 30 Example of raw and analyzed satellite imagery Source: Farmers Edge Figure 31 Example of NDVI image used to create VRT zones Source: GeoFarm Electrical Conductivity (EC) An input layer with increasing popularity is electrical conductivity mapping of fields. This method involves injecting an electrical current into the ground and measuring the conductivity of the soil. This generally relates to soil texture as small particles such as clay tend to conduct more current than larger sand and silt particles (source: Veris Technologies). Precision Farming and Variable Rate Technology 32

40 Figure 32 Example of electrical conductivity soil sensor Source: Veris Technologies Another technology uses electromagnetic waves to measure soil conductivity without contacting the soil. This method can identify salinity and soil moisture content as well. Figure 33 Example of an electromagnetic conductivity meter Source: Geonics Soil texture does not change much over time so EC mapping is a one time expense per field. It can provide valuable information on soil texture and provides another layer of information in your GIS. Target yield To know if you have achieved your goals in a precision farming system, it is important to determine beforehand what you expect for yields. There are different approaches to managing yield. The first is to maximize production of each square foot of the field by maximizing inputs. The second is to make the yield more consistent across the entire field. The overall goal of using variable rate technology is to apply the correct product at the correct place at the correct time. Profitability can be maximized by applying products where they are needed and not applying them where they are not. Precision Farming and Variable Rate Technology 33

41 Manual control It is possible to precision farm without all the fancy technology, but it is tedious and not highly efficient. An attentive operator can manually adjust their rate up or down on controllers that allow for it. One drawback of this method is there is no feedback to record whether the change was successful or not. One way manual control could be used is to increase the spray rate in heavily infested areas. Many controllers have settings for more than one rate or can be adjusted up or down on the fly. For many years, farmer s simply slowed down in heavily infested areas to effectively increase the rate. Map based control A prescription map can be developed ahead of time in the office and transferred to the in cab controller via a memory device of some sort. The controller reads the map and matches the rate applied to what the prescription map requires. Most controllers also record how much was applied where to create an as applied map. This map is useful to ensure that the product was applied at the rate expected. Figure 34 Example of map based variable rate control in sprayer Source: GeoFarm Precision Farming and Variable Rate Technology 34

42 Variable Rate Applications There are many applications that can be applied with varying rates. There are as many controllers available that can change the output of an electric over hydraulic pump, electrically driven feed rollers, mechanical gate, or pressure valve. Most of these can accept a rate input from a task controller to adjust the rate. Fertilizer Major and micronutrient can be applied at variable rates, whether the applicator is an air seeder, spinner spreader, manure injector, liquid sprayer or floater. Some of the nutrients and soil modifiers currently applied with VRT N, P, K Manure NH 3 Compost Sulphur Lime Figure 35 Example fertilizer prescription map Source: Farmers Edge Herbicide Herbicides can be variably applied based on a weed density map. The weed map could be created during the previous season s harvest if the combine operator flags weed patches while travelling the field. It could also be created by scouting the field with an ATV and marking the weed patches. Real time sensors are also available that can detect plant material and turn individual sprayer nozzles on and off when a weed is encountered. Precision Farming and Variable Rate Technology 35

43 Fungicide It is usually difficult to tell if a fungicide application is warranted since disease pressure will normally vary across the field. By scouting the field or using aerial or satellite imagery to determine the extent of disease, a variable application of fungicide can be done based on a variable rate map. Insecticide Insect infestations also normally vary across a field. By scouting where the insects are located, a variable application of insecticide can be done. It is generally recommended to do a minimal rate application for the entire field and increase the rate where infestations are heavier. Seed rate Plant populations can be optimized to make best use of available moisture and deal with problem areas such as salinity. This can create better germination and provide more competition, making it harder for weeds to get established. Hybrid varieties can also be seeded to deal with different soil conditions, topography, etc. Seeders with multiple tanks and variable drives can allow this to happen by turning off one tank and turning on another based on a prescription map. Desiccation Some areas of the field will fully mature on their own while others may be late in maturing. This uneven drying can cause problems at harvest time with some of the crop being overripe while other areas are still green. Using imagery, scouting, or non contact sensors, desiccant can be variably applied to evenly dry the crop and provide better harvesting conditions. Irrigation With water becoming a more scarce and managed commodity, better management is required. Most fields and crops do not have the same water requirements due to differences in soil type, topography, and crop type within the same field. Irrigation systems have been developed that can apply the correct amount of water where it is needed. Similar to variable rate application for pesticides, the center pivot can adjust its travel speed and turn individual sprinklers or sections on or off based on management zone maps. GPS is typically used to determine the location of the pivot in the field. Precision Farming and Variable Rate Technology 36

44 Figure 36 Example of variable rate irrigation Source: NESPAL Precision Farming and Variable Rate Technology 37

45 Yield Monitoring Managing yield data If yield data is collected every second throughout a field, imagine how many points of data are collected for your entire farm. The amount of data is mind boggling but today s computing technology is very capable of dealing with the enormous amount of data generated by precision farming. Managing this data requires a plan to develop a systematic approach to collecting, storing, analyzing and backing up your data. If this task is too tedious, you may want to involve a consultant to help you. There are off site internet based storage websites that you can back up your data to for nominal cost. It is important that you keep your data safe. If your computer fails or you otherwise lose your data, you can t get it back unless it is backed up somewhere. Once a plan to deal with the data has been devised, collect the data from the yield monitor, back it up and work with the yield data in the office. Ensure that the original raw data is securely stored in case something happens to the data you are editing. Accuracy and errors The data collected from the yield monitor will only be as accurate as the calibration and the care taken in collecting the data. Sources of error in yield data: Header cut out switch not functioning properly Header not lifted enough on row ends to stop recording Throughput lag time not entered correctly Monitor not calibrated properly Partial swaths not recorded correctly GPS errors Loss of differential corrections Combine breakdowns Sudden stops as when unloading or plugged while the threshing mechanism continues to empty Combine filling and emptying of the threshing mechanism at row ends Naming fields with the same name or not changing field names Precision Farming and Variable Rate Technology 38

46 These errors need to be filtered from the yield data to have an accurate and useable yield map. If management decisions are being made based on yield maps, it is important that the data is reliable. Figure 37 Example of yield data before and after cleaning data Source: GeoFarm The yield maps above illustrate how the map changes before and after processing. There are programs available to process the yield data, such as Yield Editor by the USDA Agricultural Research Service, available at and search for Yield Editor. The mapping software purchased or supplied with your monitor may also perform some of these functions. Calibration It is critical to calibrate the yield monitor to get accurate yield data. At a minimum, it is important to follow the manufacturer s recommended calibration procedure at the start of the season. Ensure that the pressure plate or infrared sensors are cleared of buildup. It is important not to calibrate in the middle of a field as the yield values from there on will be different and it will be difficult to produce a reliable yield map. The monitor can be calibrated against a scale ticket or weigh wagon reading. If the monitor was not calibrated, the relative yields may still be accurate when compared but the total yield may not. It may be possible to post calibrate and adjust the yields after harvest against scale tickets. Consultants working with the data will usually want the raw uncorrected data for creating their maps. Grain throughput lag time There is a time delay from when the grain enters the combine, travels through the threshing and separating mechanisms, augers and conveyers to the location that the yield is measured, usually at the top of the clean grain elevator. This is normally in the order of several seconds. To create an accurate yield map, the measured yield needs to Precision Farming and Variable Rate Technology 39

47 be tied to the coordinates when it entered the header several seconds before. Yield monitors have a calibration for this offset which is typically measured by timing how long it takes from the moment grain enters the header to when it starts appearing in the grain tank. It is important to make this calibration to have accurate yield maps. Weeds and chaff The impact plate of a yield monitor measures the entire mass of the material hitting it, whether it be grain, chaff, insects, weed seeds, etc. It does not differentiate between clean grain and unwanted material. This could possibly give false yield readings if a weed infested area of the field were harvested and entered the grain tank as a dirty sample. That is one of the reasons it is important to ground truth the field to ensure the data is reliable. Position errors There can also be errors in GPS position caused by high HDOP values, loss of GPS or differential signal, and blunders in position. If the combine stops for any reason, such as to unload or is plugged, the grain that is already in the threshing area will continue in to the clean grain tank. The position where this grain was collected will be in error since the combine is no longer moving. The position when the grain is sensed is normally moved back several seconds to estimate where it was collected. There can also be errors from not raising the header high enough to turn off the recording switch. If the yield values do not match up from one pass to the next, the error could be from position shifting of where the yield was recorded and where it was harvested. Precision Farming and Variable Rate Technology 40

48 Remote Sensing Remote sensing refers to any non contact method of getting information about a particular aspect of a field. This can mean plant health, soil quality, topography, weed pressure or anything that can be measured without contact, hence the remote part meaning operating from a distance and sensing meaning detecting physical phenomena. This is different than physically taking a soil or plant sample for analysis. Types of remote sensing There are many non contact methods of gathering information from a field. They can be as far away as a satellite or as close as a sensor attached to equipment in the field that looks at individual plants. Satellite imagery There are several providers of satellite imagery that cover Western Canada on a regular basis. One provision of satellite data is that cloud cover and darkness affect the image. Even though the satellite may have passed overhead, the image may not be usable if it is cloudy or dark. The timing of receiving satellite imagery has improved over the past few years. It is now possible to get images from a satellite pass within two or three days. This type of information can be useful to the grower if, for example, a decision on applying a fungicide needs to be made quickly. Where can I get imagery? There are several satellite imaging companies that resell their product through local distributors or on their website. Look at the list of Remote Sensing and Imagery websites at the end of this manual for some of the providers. The resolution, or size that each pixel in the image represents on the ground, varies from 30 m down to submeter level. While very high resolution imagery is available to the military via satellites, it is generally limited to 0.5 m resolution for the public unless permission is granted for higher resolution. The cost generally goes up with the resolution as the data files get much larger for higher resolution and the satellite takes images of a smaller swath on ground when the resolution is higher, requiring more passes to cover the same ground area as a lower resolution image. Aerial Aerial imagery refers to photos or digital imagery captured from an overhead elevation where the sensing device is not supported by a ground based structure. It can be from a fixed wing aircraft, helicopter, balloon, blimp, drone, kite or parachute. Cloud cover, darkness and shadows can affect the quality of aerial imagery. Aerial imagery can be Precision Farming and Variable Rate Technology 41

49 used for highly detailed images down to centimeter resolution. They can be as simple as photos taken from a remote control plane with a digital camera or as complex as a hyperspectral imaging device mounted on the belly of an aircraft. Aerial imagery has the benefit of being an on demand service and timely imagery can be quickly acquired when needed. It can also be higher resolution than satellite data. Figure 38 Remote sensing unit on a fixed wing aircraft Source: GeoFarm Non contact sensors Remote sensing can also be done from ground level with non contact sensors that measure reflected wavelengths from plants or soil to measure several properties. These sensors can be used for mapping of plant condition for later treatment or mounted on a sprayer or toolbar for on the fly fertilizer or herbicide treatment. The models which emit their own light source are not affected by cloud or darkness so they can be effective in most conditions. Google Earth as a data source Figure 39 Examples of non contact crop sensors Source: NTech Industries, AgLeader Technology Images can be imported into Google Earth and basic GIS functions can be performed on the images such as creating field boundaries, measuring areas, etc. Satellite imagery from various dates can be retrieved from Google Earth and used to compare visible variations from one time to another. Precision Farming and Variable Rate Technology 42

50 Spectral Bands: Visible, NDVI, Multi spectral, hyperspectral The electromagnetic energy that is measuring in remote sensing has many wavelengths. There are those we can see in the visible spectrum and many more we can t see in the infrared, near infrared (NIR) and many other wavelengths. Visible The visible spectrum represents a very small portion of the overall electromagnetic spectrum from about 0.4 µm to 0.7 µm. This represents the spectrum we can actually see and is associated with colors. Figure 40 Visible spectrum Source: Canada Centre for Remote Sensing For remote sensing in the visible wavelengths, the primary colors red, blue, and green are most commonly used to produce the output, which is the same as what computer displays and televisions show. A satellite or aerial image in the visible spectrum will look the same as if you were actually observing the ground from the airplane seat. Field features and variability of plant and soil color can be easily identified. Infrared, near infrared, false color infrared Plant health, soil moisture and wet areas can be detected in the infrared spectrum, ranging from 0.7 µm to 100 µm. The area of interest in agriculture is typically the near infrared which has wavelengths in the 0.7 µm to 0.9 µm range. Precision Farming and Variable Rate Technology 43

51 Figure 41 Infrared spectrum Source: Canada Centre for Remote Sensing Since we cannot see wavelengths in the infrared band with our eyes, a color transformation is used to color the image in shades that represent the degree of plant health. Bright red usually represents vigorous growth while darker red is lower growth. NDVI NDVI stands for Normalized Difference Vegetative Index and is a calculation based on spectral bands in the visible and near infrared. It is based on the premise that vigorously growing plants strongly absorb visible wavelengths and strongly reflect near infrared wavelengths. The more vigorously a plant is growing, the higher the index will be and areas with no plant growth will be lower. The NDVI image is colored to show the relative index values. Usually green is used to represent healthy plants and gold to represent lower growth. Figure 42 Example of an NDVI field image Source: GeoFarm Precision Farming and Variable Rate Technology 44

52 Multispectral Multispectral imagery refers to collecting more than one wavelength of electromagnetic imagery at the same time, generally up to 10 specific wavelengths. NDVI is multispectral since it uses the red, green, blue and near infrared wavelengths. Landsat is an example of multispectral satellite imagery. Hyperspectral Hyperspectral imagery refers to collecting and processing multiple wavelengths of electromagnetic energy, often hundreds or thousands of wavelength bands. It can provide information that cannot be seen on visible or infrared imagery. Hyperspectral imagery can be acquired by satellite or airborne sensors. One of the benefits is that the entire spectrum is collected at each point and the wavelengths of interest can be post processed. It can be used to look at crop health, leaf area index (LAI), plant canopy moisture, nitrogen and chlorophyll. It is relatively expensive, and requires significant data processing and storage, but as with most technologies, cost comes down while performance goes up over time. Figure 43 Example of Hyperspectral Remote Sensing Source: White Canada Centre for Remote Sensing Precision Farming and Variable Rate Technology 45

53 Resolution How resolution affects data quality The resolution of the data can affect decisions made using the data. For a 30 m resolution satellite image, each point of data, called a pixel, represents the average reading of a 30 m X 30 m square on the ground. If the image has 1 m resolution, each pixel represents 1 m x 1 m on the ground. Having too coarse of data resolution and too fine resolution each has its benefits and drawbacks. Inclusions in the field such as small wet areas may be missed in a 30 m image. VRT prescriptions based on the 30 m data may miss these features and alter the prescription. One meter resolution data may be too fine and make it difficult to come up with averages for zones because of too much variability in a single zone. The resolution is similar to High Definition (HD) television versus standard definition (SD). HD looks great but it highlights imperfections in the actors appearance instead of smoothing it out. Also, the amount of data increases exponentially with the higher resolution. For each data point, a 1 m resolution file will have 900 times the data of a 30 m resolution file (30 points by 30 points = 900 points). Data storage capacity has greatly increased in the past few years while costs have decreased so this is not as much an issue. Figure 44 Example of 30 m, 10 m and 2.4 m resolution satellite imagery Source: NCRST E The above example shows the same land area at different resolutions of 30 m Landsat ETM+, 10 m ATLAS, and 2.4 m QuickBird. Notice how it becomes easier to identify features as the resolution increases. Precision Farming and Variable Rate Technology 46

54 Resolution Classes There are several categories of resolution that imagery can fall into. Depending on the use of the image, different resolutions will suffice for different purposes. Submeter This is the highest resolution available. It has more use in municipal and military applications such as mapping utilities or locating targets. Satellite imagery of this resolution is more expensive to acquire as the satellite takes more passes to cover the same area and produces large amounts of data. GeoEye and Digital Globe are examples of submeter satellite imagery services. In aerial imagery, submeter is more common. 1 3 meter The next class is 1 3 meters and includes most of the aerial imagery and much of the satellite imagery used in agriculture. It has sufficient resolution to distinguish features and variability in the field. QuickBird is an example of a 2.4 m resolution multispectral satellite. Over 3 meter Satellite imagery with 3 meter or lower resolution is often used for precision agriculture. It may be more widely available at a lower cost than the higher resolution imagery. SPOT and RapidEye are examples of satellite imagery with over 3 m resolution. Ground truthing Ground truthing refers to verifying the data being collected corresponds to the true values of the readings on the ground. Ground truthing involves going out to the physical field being analyzed and looking at things like plant vigor on a certain point on an image and plants in the field at the same location to ensure the image matches the true values in the field. If imagery is not ground truthed and is classified into zones believing the values are correct, there could be errors in the zone map generated from that image Precision Farming and Variable Rate Technology 47

55 Economics of Precision Agriculture Guidance One of the quickest paybacks in Precision Agriculture comes from the addition of GPS guidance to equipment. Automatic section shutoff controls reduce waste of expensive inputs and often pay for themselves in less than one season. Payback through reduced overlap, less waste, less fatigue GPS guidance can reduce the amount of overlap between machine passes, which reduces waste of expensive inputs, allows increased yield through proper row spacing, and reduces the number of passes in a field, saving fuel, time and machinery wear and tear. GPS guidance allows a less skilled operator to still do a professional job and ensure complete coverage of a field. Autosteering takes this one step further as the operator needs only to monitor the equipment and turn at row ends. In some cases the autosteering even does this automatically. Skilled labor is difficult to attain and retain so autosteering allows less skilled operators to do a professional job. Operation in poor conditions dark and dusty Guidance systems allow for operation when conditions are less than optimal, for instance in dusty, foggy or dark conditions when normal operation would be difficult with traditional methods. Round the clock operations to maximize machine productivity Maximizing the utilization of machinery reduces the operating cost per acre as more acres can be covered in fewer days than if operating in daylight hours only. It could allow smaller equipment to cover more acres or allow larger machinery to complete the same number of acres in fewer days. This can be critical at seeding and harvest time if conditions are favorable and operators are available. VRT The economics of VRT must be considered from at least two perspectives: input and output. An input focus deals with reducing the cost of inputs (fertilizer, pesticides, and seed) by using them more efficiently. The goal is to apply inputs where they are needed, and to reduce or eliminate them where they re not needed. An output focus looks at the yield side: trying to produce the best yield possible per unit of input applied. In practice, the key nutrient is usually nitrogen (N), and the others are balanced using nitrogen application rates, target yields and soil test information. Precision Farming and Variable Rate Technology 48

56 A partial budget is the best way of measuring the economic effect of making a change, and it fits very well with the idea of an input and an output focus. A partial budget compares the old or standard way of doing things with the new and improved version. It considers increased costs, decreased revenues, increased revenues and decreased costs to calculate the overall effect of making the proposed change: in this case, adopting VRT technology. (Darling) Table 4 The Classic Partial Budget Format Source: Darling Disadvantages: Added expense: Reduced revenues: Advantages: Added revenues: Reduced Expense: Total Disadvantage Total Advantage Net Economic Advantage (+ / ) Data Requirements One of the paradoxes of modern computer based technology is that it provides lots of data but often not very much information. However, you can get started on your analysis with just the basics. Here s where the idea of focusing on both inputs and outputs comes back into play. Inputs For a simple analysis of fertilizer use and a comparison between VRT and conventional application, all you need are the nutrient application rates for each zone, number of acres in each zone, costs per lb of nutrients, and the application rates that would have been used under constant rate application. Analyzing this data can tell you: Fertilizer cost breakdown by nutrient in each field and zone Comparison of costs of VRT application (actual) and conventional (recommended). Comparison of amount of nutrients supplied (VRT vs. conventional) for each zone, field and farm. Identification (and amounts) of nutrients that would have been oversupplied or undersupplied. This could show an environmental benefit. Outputs With the addition of yield information from each zone and a conventionally fertilized check strip, you can use a partial budget to find out if VRT was profitable on your farm (at least for this field for this year). Of course, the story won t be complete until you ve considered all of the added and reduced revenues and expenses, and then repeated the analysis for several more years. Precision Farming and Variable Rate Technology 49

57 Table 5 Example Input Data Format Source: Darling Constant VRT (lbs nutrient applied /acre) VRT Yield Constant (lbs nutrient applied / acre) Yield Zone Acres n p k s n p k s nutrient cost ($ per lb) Value of crop ($ per bu) 5.15 Payback through better management The purpose of getting into VRT should be to make you a better manager. What you do with the information generated is really the key. All the information in the world will not help if it is not applied to making profitable decisions. Overall farm management As an overall farm management tool, you will be able to see which fields are most profitable and which are least. A question to ask: should you sell off non profitable land and buy or rent more land in higher profit areas? Field level management As a tool for dealing with field level variability, precision agriculture can help make more efficient use of inputs on a field by field basis. It can help make decisions on applying the correct amount of input at the right place at the right time. If each field s profitability can be maximized while promoting environmental stewardship, your precision farming program will be successful. More consistent yields One of the goals of precision agriculture can be to even out yield variability within a field. This may mean transferring inputs from areas that don t require them to areas where they will be better utilized. The cost of inputs may remain the same but overall production could be increased. Maximizing profitability If the goal of adopting precision agriculture is to maximize yields then inputs will be applied where they are expected to have the biggest returns. More inputs may be used Precision Farming and Variable Rate Technology 50

58 overall on the field, but the returns may be higher. The $/acre and $/bushel returns are expected to be greater than a flat rate approach. Profit maps When considering the profitability of your management decisions, a profit map can be used to give a good indication if you are seeing returns. A profit map for a field involves taking the sum of the outputs (yield) less the cost of all of the inputs (seed, fertilizer, pesticides, fuel, machinery costs, etc.). Within the field, this is a bit more involved and can be done in a GIS. If you have as applied maps for pesticides, seeding and fertilizer application, a cost per unit can be assigned to each point in the field. The yield map can also be assigned a dollar value per unit harvested. By subtracting the inputs from the output, a profitability map for each area of the field can be created. Look at the $/bushel and $/ac cost when considering the additional cost of VRT. If you increase the amount of crop harvested for the same inputs, the cost per bushel will be lower, even if the cost per acre is the same for that field. Look at net returns and not just the costs. Figure 45 Example of a profit map Source: Alberta Agriculture More efficient use of inputs Applying the correct input at the right place at the right time makes your farm operation more efficient. Applying blanket rates of fertilizer where some areas could have used less and some could have used more is not the best use of inputs. VRT attempts to make the most efficient use of inputs. What if scenarios: what to grow where Can precision agriculture be used as a decision making tool for deciding on which crop to grow where? This question can be asked when looking at returns for each field. Should a different crop be grown on that field or maybe a different variety would be Precision Farming and Variable Rate Technology 51