Everything Potato Growers Should Know About Remote Sensing
Have you scouted your potato crop without touching it physically?
If so, congratulations. You are a user of remote sensing.
If you haven’t, do not worry. It is a good time to start.
And if you had a bad experience with this technology, give it a second chance. Today’s remote sensing companies allow you to get the most out of their products and services.
Previously, each company had its own data format, which made communication between on-farm devices and information management tools difficult. Many producers did not know what to do with the information and couldn’t use it on their farms.
Things have improved, thanks in part to the Open Agriculture Data Alliance (OADA), a group made up of researchers from Purdue University and agricultural companies. The OADA has worked since 2014 to unify formats and improve the exchange of information between devices and tools.
Meanwhile, investors continue to rely on agricultural technology companies, which offer products that support remote sensing. According to AgFunder, entrepreneurship in agriculture and food technology received a record $10.1 billion in investments in 2017, 29 percent more than in 2016.
But what exactly is remote sensing?
Defining remote sensing
Aitazaz Farooque, assistant professor of sustainable design engineering at the University of Prince Edward Island (Canada), defines remote sensing as a non-destructive evaluation of the state of a crop.
By using a sensor and a platform, farmers can get information about a crop (and in some cases the soil), analyze the information, and compare it with reference data to make decisions about crop management.
The sensor can be a standard, thermal, or hyperspectral camera and is mounted on a platform, such as a drone, plane, satellite, or agricultural equipment.
How does remote sensing work?
Every object emits, absorbs, transmits, or reflects radiation.
Radiation is a form of energy that travels through space at different wavelengths. Together the wavelengths form the electromagnetic spectrum. But only two regions are of interest for remote sensing: visible and infrared light (near, middle, and far infrared).
This NASA video explains the electromagnetic spectrum.
To understand how remote sensing works, you have to understand the relationship between plants and radiation.
Plants capture visible light to activate photosynthesis. Near-infrared (NIR) waves do not carry enough energy to activate photosynthesis, but they do provide heat, so the plants have evolved to reflect NIR light. The intensity of the reflected light diminishes as the leaf dies. Nearby infrared sensors monitor the difference between NIR reflectance and visible reflectance, a calculation known as the normalized difference vegetation index or NDVI. A strong NDVI signal means high plant density, and a weak NDVI signal indicates problem areas in the ground (see image).
Near-infrared radiation in combination with the NDVI is the main application of remote sensing in agriculture
According to Tom McKinnon, CEO of Agribotix, NDVI images are used for various agricultural purposes. An NDVI can distinguish the areas where a crop grows well from those where it grows poorly. Using this information, the producer can apply the correct amount of fertilizer. An NDVI image can also reveal the presence of weeds, pests, and damage caused by water. This data allows the producer to identify and quantify the problems and solve them later.
What information does remote sensing offer?
As a potato producer, remote sensing can show you:
- canopy cover percentage (or amount of visible soil)
- vine nitrogen (N) concentration (or how green the plants are)
You cannot predict tuber yield or give estimated biomass (among other things) because the sensors do not “see” the tubers under the ground.
This conclusion was presented by Ann Smith and Brian Bohman, researchers from Agriculture and Agri-Food Canada and the University of Minnesota, respectively, during their participation in the Remote Sensing for Potato Production webinar, held on May 16, 2018.
If you expected more from remote sensing, do not be discouraged. Thanks to this technology, you can receive unique information on one of the most critical inputs in potato production.
Nitrogen is a fundamental element for the growth and development of plants. It is the main component of chlorophyll in leaves; chlorophyll levels affect leaf area, leaf weight, plant size, and transpiration rate.
Nitrogen deficiency can affect the quality of the plant and its productivity.
But too much nitrogen is not good either because it causes toxicity, which leads to a delay in the growth and, therefore, a plant of low quality. Besides, overfertilization results in additional costs and environmental risks.
In addition to providing data about canopy cover and nitrogen concentration, remote sensing has the potential to give you more information. However, according to the 2017 Remote Sensing Conference held at the University of Wisconsin-Madison, researchers are still experimenting with the methods and reporting.
What platforms and sensors do remote sensing use?
(The following information is based on Craig Polling’s presentation made during the remote sensing conference mentioned above.)
You can use the following platforms to know the canopy cover percentage and the vine nitrogen concentration of your crop:
- Manned/unmanned ground vehicles
- Multi-rotor UAVs (unmanned aerial vehicles)
- Small fixed-wing UAVs
- Manned aircraft
Sensors can be grouped according to their enabling technology :
- ground sensors
- aerial sensors
- satellite sensors
Knowing which platform and sensor to choose depends on:
- The physical traits to be measured
- The frequency of measurements
- Required spatial resolution and spectral bands
- Required accuracy and confidence level
- Cost of the system
Below you will find an explanatory video of each platform, the sensors it uses, the information it offers, and some observations.
The following videos and texts are meant to help you understand the technology; I am NOT recommending any company!
SPAD (Soil Plant Analysis Development), CropScan, CropCircle, GreenSeeker (video), Apogee MC-100, Handheld Spectrometers, LAI-2200, ACCUPAR LP-80, etc.
Information on a single plant or leaf.
- Easy to use
- A person takes every measurement; very labour intensive
- Small sample sizes and prone to human error
- Indicates NDVI values
- Operates day and night; with fog or cloudy skies
Manned/unmanned ground vehicles
CropCircle, GreenSeeker (video), Yara N-Sensor, Force-A Multiplex, etc.
Local canopy measurements (average over many plants).
- Easy to use
- Measurements taken continuously
- Requires driving over or alongside sensing area
- Not ideal for frequent measurements
- Operates day and night; with fog or cloudy skies
Sensors and Product
Similar to multi-rotor UAVs and small fixed-wing UAVs
- Few or no off-the-shelf products for agriculture
- High payload capacity (many sensors at once); long endurance
- Some sensors and data products need a motion or multiple viewpoints; not possible if tethered
- Difficult to manage and poor in wind
3DR, DJI, Parrot, etc.
Multispectral sensors: Sentek GEMS (video), RedEdge, Sequoia, Tetracam, Sentera, SlantRange, standard and modified RGB cameras (Red, Green, and Blue), FLIR, etc.
Hyperspectral sensors: Headwall Micro-Hyperspec, Resonon Pika, Specim, Imec, Neo Hyspex, Rikola, Bodkin, Bayspec, Pixelteq, Cubert, etc.
Multispectral sensors are a broadsword that collects less intensive spectral data. Hyperspectral sensors are like a scalpel that allows you to dissect exactly what is happening within very narrow bands of spectral content.
Hyperspectral gives you better capabilities to see what you can’t see with multispectral. It is for projects where you need discrimination power like identifying a specific disease impacting your crop and assessing severity levels
Projects that involve hundreds of acres are ideal opportunities for a sampling strategy that uses multispectral and hyperspectral sensing. You can examine the entire project area using a multispectral sensor and then use the hyperspectral sensor to get a closer inspection on any areas that look different.
Jason San Souci, GISP, Remote Sensing Scientist and GIS Strategy Expert, PrecisionHawk.
Orthomosaic maps (a composite image of a large area that is created from individual photos stitched together), digital elevation models (DEM), NDVI images, thermal images, etc.
- Easy to manipulate
- High-resolution mapping
- Low cost
- Good for frequent monitoring of medium-sized areas
- Plot-scale or plant-scale analysis, depending on sensor and configuration
- Vertical takeoff and landing
- Plan flights without considering the wind
Small Fixed-wing UAVs
Hawkeye Systems, AgEagle, SenseFly, PrecisionHawk, HoneyComb, Sentera, Trimble, Parrot (video), etc.
Sensors and products
Same as for small multi-rotor UAVs
- Longer endurance than multi-rotor UAVs, but often lower-resolution results
- Difficult to keep within the operator’s eyesight (requirement of the Federal Aviation Administration [FAA])
- Usually more expensive than multi-rotor UAVs
- Good for monitoring larger areas
- More difficult to use than multi-rotor UAVs
- Manual takeoff and landing
- Must be aware of wind when planning a flight
- Small planes equipped with high-resolution RGB/NIR or modified RGB cameras for NDVI
- Usually provided as a service
- Can cover large areas at lower resolution
- Somewhat expensive
- Not well suited for frequent monitoring
Author’s Note: TerrAvion makes me a necessary clarification about manned aircraft platform. Read it at the end of this article (Responses section)
Satellites — Continuous Coverage Constellations
- Many small satellites
- Every point on Earth is imaged every 24 hours
- Low-resolution, often poor quality imagery, RGB or RGB+NIR only
Satellites — Tasking Constellations
- Few satellites in a constellation
- Can revisit frequently, but limited coverage
- Must be commanded to image a given area
- Better sensors; up to 28-band multi-spec, high-resolution visible and near-infrared (VNIR; 30 cm of ground sample distance [GSD])
What are the benefits of remote sensing?
Some of the benefits of remote sensing are:
- Facilitates the general monitoring of the crop (what is happening right now)
- Decreases the possibility of nitrogenous overfertilization
- Speeds decision making
- Reduces overfertilization, leading to less nitrate pollution in water
- Improves irrigation management
- Optimizes the use of supplies
- Increases crop yield
- Attenuates losses in production and quality
Read the advantages and disadvantages of remote sensing, formulated by GrindGIS.
Drones or satellites: Which is more useful for producers?
Ann Smith and Brian Bohman agree that satellites have lower demand than drones because satellites have difficulty in obtaining information on cloudy days. Satellites also do not offer enough resolution to see details at the crop level (see webinar).
But they say that gaining information through the use of drones is more expensive, and although the situation is changing, producers have had problems interpreting data and applying it to their farms.
According to the researchers, the company Planet may be tilting the balance. A complaint about satellite images is that they are not taken frequently enough for the producers to solve the problems on time. The daily images offered by Planet would change the rules of the game in the agricultural digital space.
A medium-term solution is platforms that combine several types of data. An example is watchITgrow. Using satellite and drone images, weather and soil data, and yield prediction models, watchITgrow helps potato growers in Belgium improve their farms.
In the end, the researchers conclude, producers will select the platform that provides the best information at the lowest cost.
The following table summarizes the features of Multi-rotor UAVs (Quadcopter), Small fixed-wing UAVs, Manned Aircraft, and Satellites.
Are there cost-benefit studies on using remote sensing in agriculture?
At the end of the webinar, Ann Smith said because studies done to date have been based on different methodologies, it is not possible to say whether the investment in remote sensing is profitable.
Based on a recent study by the U.S. Department of Agriculture, National Geographic argues that “Returns are estimated in dollars per acre, so larger farms stand to gain more from the technology and can allocate a larger budget to precision agriculture.”
The study was conducted on precision agriculture, not on remote sensing. But because the latter is included in precision agriculture, the information is valid for this article.
Watch the promotional video of that study.
“Satellite and aerial imagery are on the way to be widely adopted by farmers.” That was said by Ivanov Igor, co-founder of Gamaya, a company that offers solutions based on hyperspectral images and artificial intelligence. His testimony is based on a study by the University of Nebraska-Lincoln.
Another investigation carried out by Goldman Sachs and cited by the same author concludes, “The technology solutions that are offered and will be offered in the future increase the value of land by making it possible to produce significantly more crops per acre.”
These are important data if we take into account the U.N. prediction that in 2050 the world population is projected to reach 9.8 billion (the current world population is 7.6 billion). This means we will have 2.2 billion more mouths to feed.
As a producer, you have at your disposal a variety of technological solutions that allow you to “know” your crop. And you can do it at your own pace and according to your budget.
But there are still bottlenecks that must be solved.
For example, many companies offer what they think producers want and not what they need.
For Lisa Prassack, president of PrassackAdvisors, access to images has never been easier or had more options. She highlights the interest of producers to use them to get the most benefit. But what is the use, she asks, for a producer to say, “I have this beautiful image of my land, but how do I turn it into information that improves the efficiency and/or profitability of my farm?”
If these limitations are not corrected soon, we could end up asking ourselves, like Ehud Olmert, What is the logic of investing money in agriculture if it does not work?
Do you use remote sensing? What has your experience been using this technology? I’d love to hear your thoughts. Send them to firstname.lastname@example.org
About the Author
Jorge Luis Alonso G. is a freelance B2B technology content marketing writer. He writes white papers, articles, and blog posts about agricultural innovations strategies, including precision agriculture, remote sensing, big data, IoT, and food technology. He lives in Argentina.