Interesting factoid.. it takes more water to produce a kg of cheese than it does to produce a kg of chicken meat or pork (presumably even less for the pastured chicken and pork that I buy, since it includes the grains that are typically fed to conventional chickens & pork).  This can't make the vegetarians happy.. I should decrease my cheese intake and increase the meat in order to conserve water?

None of these problems will be solved in the pure consumer model.

Production-consumption is the answer.

Solar cells, make hydrogen, emit water, collect water, drink water.

Every split-level needs to be it's own terrarium.

If I flush my toilet here in the Northern Sacramento watershed every third gallon will end up headed south towards a golf course in the LA basin. Flush that toilet in San Francisco and every gallon goes into the Pacific after treatment. In both cases the water should have passed through the Sacramento river delta without human interference. Locality matters here.

A pasture grazed milk cow in damp Vermont is going to require a fraction of the water input that a confined CAFO'd milk cow in Bakersfield CA.. A completely different water requirement comes from a riparian grazed cow in Costa Rica. Assigning a fixed quantity of water used to a liter of milk or a kilo of beef is just meaningless propaganda.

Plus some of the numbers are just goofy. The stated water consumption of a Kg of millet is 5000 liters when millet is a semi-arid crop almost always grown without irrigation. As a scientific statement of how much water passes through the millet plants that's accurate but as a statement of "human water use" it's totally useless. There's no irrigation percentage and runoff depletion figures.

Finally, since I didn't immediately see a reference to the water wasted in the production of the common cotton t-shirt I suspect we're being suckered. The Aral Sea disaster is the worst example of environmental damage due to water diversion and that was entirely done to produce cotton. Instead the bad water user examples are beef and milk which have highly variable water use.

I think we see here another shot in the vegan netwar.

Put the Carbon Back

Cheese produced from naturally-rainfed pastures in Vermont or Wisconsin makes a lot of water sense; cheese produced from confined cow-feeding operations near Phoenix, Arizona does not.

The vast majority of water diverted for primarily for non-consumptive uses (e.g. dishwashing, bathing) eventually returns to the source river or watershed and is much less environmentally consequential than water that is diverted and then permanently lost to evapotranspiration (e.g., raising irrigated rice or cotton in the middle of a desert).

There are profound differences between these situations.  Indeed, we all need to be more aware of our water uses and the impacts they have.  But not all water uses are created equal!