Data As Key Enabler to Regenerative Agriculture

Regenerative agriculture is here to turn farming practices more nature-friendly and sustainable. It questions traditional practices that have taken a toll on the environment – despite increasing productivity at significant levels over the last few decades –with substantial greenhouse gas (GHG) emissions that contribute to global warming and climate change. According to the World Resources Institute (WRI), in 2020, the agriculture sector was estimated to be responsible for around 13 percent of global GHG emissions, excluding land-use change and forestry. The challenge is that climate change has an impact on agriculture itself, as crops are dependent on specific climate conditions to thrive. Pulling the string further creates a domino effect that affects the whole food chain.

So, what exactly is regenerative agriculture? There's still lots of debate around the topic, with no consensus on a definition that shall continue to evolve. What is clear, though, is that it should aim at reducing climate impact, improving soil health, promoting biodiversity, and making better use of natural resources, all of that while also creating a system where farmers can prosper; after all, such changes require a huge investment that they can’t afford alone – public and private investments would be needed.

Food-chain companies are now in a quest to understand how they can effectively foster the adoption of new practices in the field. Here’s where things become exciting for data professionals like me, as it is imperative that they measure how much each step on the way – from plant to farm to fork – adds to the GHG equation. In other words, data must be collected at the most diverse stages to enable informed decisions, including incentives to farmers.

How can data support regenerative agriculture?

Traditionally, agricultural field trials have been the most effective way to measure the outcomes of a given product or technique applied to crop and land management. However, it is not enough to measure regenerative agriculture's impact as it does not scale. It is necessary to collect granular farm and field data at a scale that will allow farmers to measure the impact of their actions (i.e., practices) and recommend changes to them. Therefore, defining the right set of indicators that will both measure and guide farmer’s adoption of sustainable practices is key.

​Regenerative agriculture is here to turn farming practices more nature-friendly and sustainable.

 Fortunately, with more technologies supporting their day to day, farmers can now collect agricultural data in a variety of ways and for different purposes. For example, laboratory analysis and ground sensors can portray the soil health; aerial imagery (e.g., drone, airplane, and satellite) can show the biomass evolution over time and also estimate yield; tractor sensors deliver fertilizer application data; weather stations capture climate conditions, which support application decisions.

It is worth mentioning there are already successful implementations of machine learning (ML) models in critical areas. To list a few, cloud detection and removal from satellite imagery have been explored by many scientists with the objective of improving the quality and frequency of available field data; computer vision using artificial intelligence is also enabling growers to accurately estimate yield and allocate resources for harvesting; and remote sensing techniques have been applied to estimate soil quality indicators.

The real challenge is perhaps how to integrate all these data, identify and quantify how each element influences regenerative agriculture outcomes, and collectively build such knowledge. Collectively, yes, because ‘it takes a village’ to restore climate stability, requiring several actors in the agriculture and food industries to work together in defining ontologies and standards that will facilitate the interchange of data and information. These are necessary for measuring and verifying the influence regenerative practices have over GHG emissions.