Precision Agriculture

Written By Sahil Muji

State of the Farm

Agricultural production is responsible for roughly 13% of global CO2e emissions. Of the nearly 7.5 billion tons of CO2e emitted from farming annually, more than 20% comes from nitrous oxide (N2O)[1]. N2O is a potent greenhouse gas with a warming potential 290 times greater than carbon dioxide. These emissions largely come from nitrogen fertilizers which are essential to many farm operations. Farmers rely on fertilizers and other inputs to maximize their yield, but they often overapply these chemicals as they are unwilling to risk their harvest. Excess application results in emissions and chemical runoff into surrounding waterways. The application of fertilizers is only one part of the emissions from chemical additives. Production accounts for one-third of the nitrogen fertilizer lifecycle emissions, not including the impact of transporting and storing the product.[2] In the United States, nitrous oxide emissions constitute 6% of total emissions, driven by the production of oilseeds and grains which make up more than 70% of cropland.[3] Despite its warming potential, there is no consensus on how to address these emissions.

Fertilizers are not only a burden on the environment but are also a major cost for farmers. A handful of fertilizer producers control over 80% of the market, leaving individual farms at the mercy of market prices. With no way to influence the price of crucial inputs like nitrogen, farmers focus on maximizing output, typically measured in yield. However, attempts to boost production have led to overapplication of fertilizer, creating an unnecessary cost for farmers. Chemical fertilizers are a large portion of production costs, especially for grain crops like corn and wheat. In 2022, fertilizer expenses alone accounted for 36% of the average corn farmer's production costs.[4] Even though prices have moderated since their peak, they continue to hurt farm income. Farmer incomes fell by 16% in 2023 and are projected to dip another 25% in 2024, driven in part by higher production costs. Fertilizer remains an important tool in the farmer’s tool belt, but over-dependence is a major risk. Prices can see large swings in short periods, making a big difference in farm profits. To reduce their production costs and exposure to price swings, farmers need a way to cut down on their use of fertilizers. 

Potential Solutions

Alternatives like microbe-based fertilizers could help farmers reduce their use of chemical fertilizers, but they face headwinds before they can be adopted at scale. Microbial fertilizers would make nitrogen in the soil available to crops, ideally allowing farmers to cut their use of synthetic nitrogen without any loss in yield. In reality, the value of microbial fertilizer is being debated, with some researchers skeptical about its efficacy. Even if alternatives are proven, they tend to cost more than the fertilizer they replace. Farms are risk averse and many struggle to stay profitable. They will not invest in green alternatives that will lower their margins and introduce new risks to their fields. For alternatives to be a practical pathway to emission reduction, farmers need to see a financial benefit to switching. Farmers have no reason to take on additional risk unless they think they will see a return on their investment.

Farmers, however, can lower their environmental impact without switching to alternatives or sacrificing productivity by optimizing their use of chemical inputs. Many farmers over-apply fertilizers, leading to significant nitrogen runoff. Nitrogen Use Efficiency (NUE), a measure of how much nitrogen is used by crops, helps demonstrate how much nitrogen goes to waste. In the United States NUE hovers around 60%, resulting in a staggering 3.5 million metric tons of nitrogen needlessly lost to the environment. [5] Globally, NUE stands even lower at 40%, indicating substantial room for improvement. With some help, farmers can curtail their use of these chemicals without any loss in productivity or yield. This helps farmers lower their environmental impact and improve profitability by cutting down on wasted resources.

However, farms cannot just cut their application rates across the board. Farmers have applied fertilizer for decades for a reason.  Making uninformed blanket decisions on application rates could lead to disaster. Reductions in fertilizer application need to be informed by data on soil and crop characteristics. Access to data can help farmers gauge the impacts of the decisions and make changes to better manage resources.

Precision agriculture (PA) can help farmers gather information on their fields and crops to better inform decisions on chemical use. The goal of PA is to use new technologies to deliver individualized treatment to crops. Crops from across the same field often have different needs based on variations in soil, elevation, and a myriad of other factors. GPS and sensors are already used to capture high-level data from across a field and show discrepancies between crops. Equipped with this data, farmers can use variable rate technology (VRT), another facet of PA, to target the application of inputs like fertilizers and pesticides. By providing each crop with what it needs farmers can cut down on production costs and better manage their land. Technologies like VRT and GPS are already prevalent on farms, with the majority of new equipment already incorporating these innovations. The availability of useful and accurate data currently limits the impact of these tools. Higher quality and more accessible data can help farmers make the best decisions for their crops and their bottom line.

The Opportunity

Today, most farms that use VRT rely on yield monitors and maps to guide their application of fertilizer. A yield monitor is used to capture data on yield at the time of harvest. This info is used alongside GPS data from the harvester to create yield maps (figure 3). Equipped with a yield map and information on previous operations, farmers can prescribe different treatments across a field based on the measured productivity. These maps provide a good starting point, but they have limits. Yield maps can only be made once a season, which makes it difficult to capture short-term changes. With data collected at more regular intervals, farmers would better understand the immediate impact of their decisions and could make changes accordingly. They also only capture data on yield, which leaves the farmer to figure out how to best adjust their practices. Farmers need better data to maximize the impact of VRT.

Soil maps can help provide details that yield maps do not capture. These maps can capture soil characteristics such as organic content, pH, nutrient levels, and more. They show the outcome of historic soil management practices with more precision than yield maps. Depending on the technology used they can provide varying levels of detail. These data points can help farmers understand the impact of their treatments with greater accuracy. But these maps are still not perfect. Traditional soil tests are conducted by taking soil samples from across the field and sending them to a lab to be analyzed. The whole process can take weeks, which makes it difficult to quickly act on the information. These maps are also not as precise, as each sample could represent up to 10 acres. This does not provide enough detail for farmers trying to decide on application rates for fertilizer.  

For farmers to earn the greatest benefit from PA, they need data-gathering solutions that are easy to deploy and use. Part of the reason that yield sensors and VRT have become so prevalent on large farms today is that farmers can deploy them into their existing operations without significant training. These plug-and-play technologies remove the technical barriers that some might associate with precision agriculture, making them far more accessible to farmers. Data-gathering technologies that rely on farmers developing sophisticated technical skills will face enormous barriers to adoption. And relying on external expertise to gather data makes any platform difficult to scale. Agronomists or other experts may be capable of gathering the data a farmer needs, but at scale this would require a lot of manpower, making it too expensive for farmers.

Solutions that can be deployed without technical expertise are more likely to be adopted and function at scale. Removing the burden of data gathering from the farmer, they can focus on the insights earned from the data and how it can be used to improve operations. The startup Stenon, for example, has made data capture as easy as possible for the farmer. Their sensor technology allows farmers to analyze their soil as often as they like and get immediate results without complicated gear. Solutions that can provide similar ease of use for crop and soil data will be more approachable for farmers.

In addition to making data collection easy, the data needs to help farmers. For most farmers, data alone is not sufficient for them to change their practices.  It needs to be analyzed and interpreted before farmers can use it to make decisions on nitrogen application or other practices. Hyperspectral cameras, for example, capture hundreds of wavelengths that can be used to determine soil characteristics. This data would not be useful to farmers without some outside analysis. Deep Planet, a crop insight platform, demonstrates how analysis can distill complicated data sets from hyperspectral imaging into a simple dashboard. This dashboard is infinitely more useful to farmers than the raw data that Deep Planet collects. Any data-gathering solution needs to provide insights for farmers. Without analysis, the data provides no value to farmers and is not going to see much adoption.

For farmers that grow oilseeds and grains, the majority of cropland in the US, the cost is a major concern. These operations are huge, typically spanning thousands of acres, and operate on small margins. Even small costs can make a big difference in income when an acre of corn typically generates $200 to $300 in profit. Many of the existing data collection solutions have proven too expensive for these farms and have instead focused on vineyards or other operations that grow commodities with higher margins. Grain farmers, in general, will not adopt new technology unless they see a clear return on investment. To accomplish this, sensors need to be easy to scale on large fields and require limited manpower. Solutions need to demonstrate to farmers that the cost of acquiring data is justified by savings from reducing fertilizer usage. 

Why Now

Today many of the important precision agriculture technologies have reached maturity and are becoming mainstream. GPS receivers for tractors, once a rarity on farms, are now common fare on new pieces of equipment. Both yield mapping and VRT steadily increased in use since 2000, reaching 44% and 37% of planted corn acres in 2016, respectively.[6] And adoption of these technologies has only increased since. These technologies are the infrastructure needed for farmers to make use of data in a meaningful way. If GPS receivers and VRT were not available on farms, collecting would be useless because there would be no way for farmers to benefit from it. Now that precision agriculture technology is becoming commonplace in farming operations, there is a need for detailed and accurate data.

There have been significant technological innovations over the past decade that make data acquisition more accessible. Many sensors have become cheaper, more compact, and more user-friendly. These advancements make it possible to try new solutions that were not previously possible. For example, multispectral sensors were originally equipped on satellites to capture data on the Earth. Now sensors have gotten lighter and more compact, allowing them to be equipped on drones or grounded equipment.  Advances in AI have also improved our ability to analyze data. Once it is trained, AI would be equipped to handle the vast amounts of data that might come from 1,000+ acre farms. Algorithms can help sort through data, identify patterns, and return an output that can be used by farmers.  These technological improvements have created new options for data acquisition and interpretation.

One of the benefits of precision agriculture is that it is not contingent on any one technology or strategy. It can be adapted to new hardware, different crops, and changing production processes. For example, data collection could be used to support microbial fertilizers by identifying parts of a field that have a low microbial content. It is also compatible with regenerative agriculture, another strategy to reduce the environmental impact of farms. Crop and soil data could help farmers choose the best companion plants and validate the benefits of sustainable practices. Regardless of changes or advances within the agriculture space, having access to better data helps farmers manage their land. As the environmental impact of the agriculture industry becomes a bigger point of discussion, data will play an important part in the conversation.

Over-application of fertilizer causes nitrous oxide to be released into the atmosphere and adds to farmers’ production costs. Fortunately, there is an opportunity to help farmers by providing them with data on their fields. Data on soil characteristics provides guidance on how much fertilizer each crop needs. Equipped with this data, farmers can make decisions on how to minimize excess nitrogen being deployed without taking on any risks to yield. Data collection solutions need to be easy for farmers to use and demonstrate clear benefits to their bottom line. Effective solutions will be able to lower production costs without any losses to yield, helping farmers earn more.

Sahil Muji

Sahil is a venture fellow with the Orca Climate Fund from 2023-2024. He graduated from NYU Stern with concentrations in finance and sustainable business and has held roles in companies ranging from owner-operators of sustainable infrastructure to startups focused on plastic pollution. He is focused on finding ways to help companies to better manage their natural capital and improve biodiversity. He is currently working on the sustainability team at Abrdn, a global asset manager.

Previous

Decarbonizing the Textiles Industry