[I've returned to the world of home Internet, finally coming out of a week without, after a road crew cut the cables going to the entire village. The week was also visited by a dearth of electricity, so I'm now catching up.]
Water quality, water availability, and water rights are increasingly at risk around the world. Recently I’ve been reading up on new ideas for what to do about it, and have been especially interested in attempts to calculate ‘water footprint’ and to coax ‘virtual water’ to flow in the most efficient directions.
Water footprint formulae, like carbon footprint calculations, attempt to quantify the total amount of resource used in making, transporting, using and disposing of a particular product. You can go to waterfootprint.org and find the total amounts of water used for all kinds of things, from coffee (140 liters/cup) to clothing (2700 liters for one shirt). These calculations are difficult and imprecise, but they’re a start.
As if it’s not complicated enough to figure out the embodied water use of every product, calculating an individual’s water footprint appears to be even more challenging. Not because it’s hard to add up the total gallons used, for every process and product, but because people’s water footprints have different levels of impact depending on the availability of water in their local area.
It seems that a footprint calculation has local and global components: local availability of the resource, plus global trade in the resource. In calculating carbon footprint, I don’t think the local component is taken into account (ie the challenge of procuring petrol in India vs in the United States). But in a water footprint, it is a prominent piece. A farmer in water-rich southern China, who uses x amount for irrigation should, all else being equal, have the same water footprint as a farmer in drought-ridden sub-Saharan Africa, but the impact is less. (Water footprint)-(Water availability)=impact. Or something.
That’s water being used. Water also moves: it moves across the world in natural form as rivers, lakes and glaciers cross borders, in man-made physical form, and in products that people move across borders. Man-made physical water trade comes in the form of massive projects: China’s South-North is a huge plan to transfer water from to the water-scarce Huang-Huai-Hai (Yellow River area) Plain in the north, from the water-rich south. India’s National River Linking Program aims to build aqueducts and upwards of 25,000 km of canals, connecting 37 watersheds around the country. Globally, projects like these that are already completed transfer around 490×10^9 m^3/yr (pdf) per year. Planned projects, mainly in the Americas, Asia, and Europe, would add 1,150 x 10^9 m^3/year.
Virtual water encompasses the water footprint that moves when a product is transported. For example, a melon grown in California, transported to New York, would carry with it the water footprint of its production. That movement is ‘virtual water.’ These days, virtual water moves around the world in all manner of ways. It turns out the flows are anything but resource-efficient, with virtual water often moving from water-scarce areas to water-rich areas. Some opponents of massive water transfer projects have been pointing toward aligning virtual water flows with the direction of the needed water transfer.
A case study of virtual water trade of agricultural products around the various parts of India asks what it would take to correct these flows, and introduced me to some of the issues. Citing a 2006 study, it says:
North China exports 52×109 m3/yr of virtual water to south China, a volume which is more than the maximum proposed water transfer volume along three routes (38 – 43×109 m3/yr) in the South-North Transfer Project. The study therefore concludes that if the “perverse” direction of virtual water trade in China can be reversed, it can act as a better alternative to physical transfer of water across basins. It is with a similar logic that the idea of virtual water trade within India is being proposed as an alternative to the NRLP.
The study goes on to look at virtual water transfer in India, which mostly flows from the dry north-west areas of Punjab and Uttar Pradesh to wetter eastern areas of Bihar, and to the south, to Kerala. It’s pretty dramatic: In Haryana, Uttar Pradesh and Punjab, the main water exporters, per capita water resources 2176, 2922, and 3554 cubic meters per year respectively. In Jharkhand, Bihar and Orissa, per capita water resources are 4580, 6898, and 8710 cubic meters per year respectively.

Though it would seem logical that areas rich in a resource would export products that require that resource to poorer areas, it turns out that decades watching international trade show it’s often not the case. The study considers a bunch of other factors, trying to find one that might be driving the exports in the wrong water direction.
The two factors that seem to have the most impact on agricultural produce exports are ‘per capita cropped area (area under agricultural production)’ and ‘access to secure markets’. There are many political, cultural, and economic reasons why this might be.
Perhaps partly because in areas where there is more growing, there is more irrigation and therefore more subsidy, propping up an illusory water availability perception, or driving down water ‘prices’. ‘Per capita cropped area’ may not take into account fallow land (not that there’s much in India), or land that could be productive if given the same level of public infrastructure. Other production factors, like local knowledge, pests, branding/perception, traditions also must play a part. And, access to secure markets is a diplomatic phrase encompassing the buying policy of the Food Corporation of India, which buys a huge percentage of all crops in its massive food security and subsidization projects, favoring some areas (that happen to be more water-scarce) over others.
The conclusion is that water scarcity is not the most important factor. Yet. Until it is, it won’t drive agricultural transport. The study makes some interesting points about the main factors, land availability or access to markets:
By importing food grains from a land rich state, a land scarce region is economizing on its land use. Following the virtual water trade logic, this can be termed as virtual land trade. A land-scarce region (such as Bihar) would import crops from regions where land productivity is higher (for instance, Punjab). In order to produce the same amount of food in Bihar, Bihar would have to employ more land than Punjab (Aggarwal et al., 2000). If, and as long as, land is the critical constraining resource, Bihar would like to economize on its land use, even at the cost of inefficient or incomplete utilization of its abundant water resources.
This study didn’t include the environmental or economic costs of transporting goods, which might change the equation as well. Though that’s another set of thorny issues — trade is a good thing, so is locally produced and consumed food. There are numerous ways to look at this information as an opportunity to clean up policy, start small enterprises, and educate communities on what they can do. I’ll be looking into it more, as I am able.