Agriculture is one of the biggest human causes of greenhouse gas emissions. As well as the process itself, it’s one of the big causes of deforestation, further squeezing our ability to deal with carbon in the atmosphere. Given the population looks like it will only grow, how do we manage to cut those emissions while still feeding everyone?

In this video, Adam Formica (a student at the Oxford School of Geography and the Environment) explains how he’s been trying to help make agriculture more efficient using data and computer modelling. By doing so, we can produce less waste and use less land, leaving more space free for forests.

A lot of work, information and expertise is needed to make models like this. Adam has been lucky in his DPhil to have gotten to know a wide network of people in agriculture, from government officers to smallholder farmers. As well as gaining an understanding of how the systems work, it has also helped him to work out what data is available and where. It also gives him valuable collaborations to test different models.

One way of making these efficiency savings is by predicting crop yields (for example, Adam worked on trying to predict palm oil production in Malaysia). Luckily, agricultural producers keep a lot of historical data. When you combine this with data from satellites, machine learning models can begin to anticipate crop production. If you know what yields might be, you can make things more efficient my getting the right workers to the right field at the right time. Similar models could also aid efficiency by identifying what leads to greater yields, ensuring the maximum output from each area. So you could get more food, without needing more land.

Another angle is looking at waste. Around 30% of total food produced globally is wasted or lost, with half of that coming from supply chain inefficiencies. If the same efficiency savings from developed countries were spread globally, total waste could be reduced by about 5%. So another aspect of Adam’s work was to build simulations that model how road development and planning can lessen waste and increase yields. In Senegal, for example, he worked on a simulation to look at better delivery routes, so time-sensitive crops could arrive more quickly.

The potential land gains from making the supply chain more efficient like this is massive. Adam calculates that with 5% less waste, we could free up roughly 3.25 million square kilometres (more than 13 times the size of the UK).