I’ve been thinking about food waste a lot lately. Why? It started when I perused Drawdown’s list of top solutions to global warming. Reduced food waste is ranked third, and it’s the highest ranked solution that almost anyone can take action on. But before we can take action, we need to be informed. This post seeks to do so, empowering you to take part in the conversation that needs to happen to solve this gnarly problem!
What is food waste? The Food and Agriculture Organization of the United Nations (FAO) defines food waste or loss as a decrease in food mass, where food includes products originally intended for edible human consumption.1 The FAO further distinguishes between food loss and food waste. Food loss happens before food reaches retailers and consumers, and is largely unintended and very challenging to avoid. Food waste happens later in the supply chain, and is more easily avoidable and caused by human behavior. The chart below shows the percentage of food loss and waste at different stages in the supply chain by calorie, and for the developed and developing world. For simplicity, I use food waste to refer to loss at all these stages. You might notice that incorporating the packaging waste associated with food is outside the scope of the definition. This is obviously another big problem contributing to both climate change and plastic pollution, but I will defer the topic to another time.
You might be asking how much food is wasted. According to the FAO, fully one-third of the food produced globally for human consumption goes uneaten. This amounts to about 1.3 billion tonnes of food per year. This amount sounds staggering, but it’s hard to visualize. To make an attempt, I thought about the Rose Bowl2 and whether we could fill it with our wasted food. I did the math, and calculated that we could fill the Rose Bowl almost 13 times over with all our wasted food. But that’s not in a year, that’s every single day!
This loss occurs on the farm, during transit and processing, in stores and restaurants, and in the home. The last is where we come in. I personally have wasted food even in the past few weeks, even though I’m trying hard not to. It’s easier to waste food than you might think, until you pay attention. From the milk that goes bad to the leftovers that your child refuses to eat to the overlarge portions at restaurants, it’s nearly ubiquitous in American culture. Food is relatively inexpensive, and it’s challenging to pay attention to food waste given everything else in our lives.
What does all of this have to do with climate change? While less recognized than other climate change contributors, food waste is responsible for approximately 8% of global emissions; if it were a country, it would rank third. The most conservative Drawdown scenario estimates that 70.5 gigatons of CO2-equivalent emissions could be averted if we reduce food waste by 50%.
These emissions come from several sources, including agriculture emissions; energy for food processing, storage, transportation, and preparation; land use change and deforestation; and decaying food in landfills. Agriculture and landfill emissions are particularly troubling as they are primarily methane emissions. Methane is approximately 30-times more potent than CO2 as a heat-trapping gas (over 100 yrs). However, this also poses an opportunity: methane breaks down more quickly than CO2, which means avoiding it will help reverse global warming more quickly. Composting our food thus represents a huge and easy approach to reduce some of the impact of food waste.
However, as with many parts of climate change, food waste is a complex problem – composting is not a silver bullet. One of the most compelling visualizations I’ve found of the problem’s complexity is illustrated below in a rendition of the Global Food System. Food waste is included here but in our actual lives it’s quite hidden, one of the many externalities in our lives. Understanding cause and effect is obviously a challenge!
In data science, however, we love complex problems! Here are a few examples of techniques that deal with complexity. We have developed clever ways to represent and manipulate data. We know how to detect interacting variables, which are ubiquitous in complex systems such as the food supply chain. Next, we’ve used recommendation systems to tune results to a particular person or situation. Finally, an area that I’ve long been excited about is ensemble learning, in which we train multiple learners on the same problem and combine them when making predictions. This can help increase predictive performance when compared to individuals in the ensemble; we can think of this as similar to the Wisdom of Crowds phenomena.
I’ve been brainstorming on ways to contribute to reducing food waste using the techniques above or others, and will say more in a future post. Meanwhile, I hope you’re now inspired to be part of the solution. Have some ideas? Please leave a comment!
Footnotes
Photo courtesy of Nick Saltmarsh
1. I’ve simplified the definition given in the paper.
2. This exercise was inspired by the opening of the book “American Wasteland: How America Throws Away Nearly Half of Its Food (and What we Can Do About It), by J. Bloom”
