The Water Footprint of your Food
This was originally posted on my first blog, Data vs Food, in August 2017
At almost the same time as my first blog post Eva Murray tweeted Makeover Monday Week 30 – How Thirsty is our Food?
A food related dataset, how could I not write my second post about this?
The visualisation for this week was originally from Statista, the data itself comes from waterfootprint.org and details how much water is used in the production of our food. This volume of water can be broken down into three “types”. I wanted to highlight this split but also compare the different food types. This is a really interesting topic, I knew that producing certain foods like beef used a large amount of water, but I was stunned by just how much. Over 15,000 litres for 1kg!
If I’m ever stuck for ideas on how to progress I’ll usually have a look back at previous recaps, there’s such a huge wealth of great visualisations and commentary here. In week 8 a viz by Florian Ramseger used a treemap to compare the metrics for each country, I thought this would work well as the main part of my viz.
As tempting as it is to start scouring twitter to see what everyone else is creating, I try to stay away until I’ve got a good idea of the direction that I’m going. One of the first things I’ll do is sketch out some rough ideas and then try to draw what I think my final viz will look like.
So this is roughly what I planned to do. The top half shows my first vague ideas; the bottom left is my final plan with any actions or other notes to the right. Once I had this idea in my head of water pipes flowing down with the treemap in the centre, I stayed with it almost to the letter. I find this rarely happens when I’m sketching things out, I’ll usually come across something that doesn’t work. I’ll then literally go back to the drawing board and see what other approach might work. And here’s how it turned out.
To get the grouping within the treemap I had to pivot the three types of water to create ‘Water Type’ & ‘Water Volume’ fields, but this is nice and straightforward in Tableau. One thing I did get stuck on was the labelling. I wanted to only label the green water sections with the name of the food item. I tried a few calculations, they labelled the treemap correctly but they also changed the grouping from food items to water type. In the end I annotated the marks with the food item names, a little tedious but it worked.
Giant Couscous, Roasted Vegetables, and Feta
Serves 2 to 3
This is a straightforward recipe that is quick to cook and really tasty. My wife, Sophie came up with this one, so all credit goes to her!
1x Red Pepper
1x Red Onion
200g Cherry Tomatoes
150g Giant couscous
A small bunch of fresh Coriander
Juice of half a Lemon
Olive Oil spray
Salt & pepper
Chop up the vegetables:
Cut the red pepper into slices
Cut the courgette into batons
Cut the red onion into 8 chunks
Halve the cherry tomatoes
Put all the vegetables onto a roasting tray, season with salt & pepper, spray with olive oil, and roast for about 15 minutes in an oven preheated to 200C/180C Fan.
Put the giant couscous into a pan of boiling water and cook for about 10 minutes or according to the packet.
Drain the couscous and crumble the feta into it along with the lemon juice and a dash of oil.
Plate up the couscous and add the roasted vegetables on top.
Add some more crumbled feta (this isn’t required, but I love feta).
Chop the coriander and sprinkle over, enjoy!
When I was writing this up I had a search and found out that giant couscous isn’t actually couscous. It turns out that it’s toasted pasta balls, who knew?
And to link this back to the original dataset, I did some rough calculations of the weight of the ingredients used in this dish and calculated how much water it would have taken to produce them. I made the assumption that the giant couscous, by weight, was 50% cereals and 50% eggs. This accounted for over 40% of the water footprint.
As a comparison, here’s the water footprint of this dish vs. the same weight in beef, chicken, and pork.
It speaks for itself really.