News

Groundwater: the unknown resource

Development Policy18 Mar 2013Mark Bierkens

Non-renewable groundwater abstraction leading to groundwater depletion has global dimensions. It should therefore be addressed by a mix of solutions in cooperative projects between international and regional experts and local stakeholders.

We all know the figures: worldwide 17% of crops are irrigated, accounting for 40% of food production. What is less known is that almost half of the irrigation water required comes from groundwater and that almost 20% of the groundwater abstracted is non-renewable, i.e. it exceeds the regional natural and artificial recharge.

Non-renewable groundwater abstraction leading to groundwater depletion occurs in several hot spots around the world, with north-western India, parts of Pakistan, north-eastern China, Iran, the Arabian Peninsula and the US (the Central Valley of California, High Plain aquifer) taking the lead. Up to now, population increase and economic development leading to an extension of irrigated areas have been the main drivers for increasing groundwater abstraction and depletion, (by 200% and 300% respectively since 1960). However, several studies indicate that, during prolonged periods of drought, groundwater abstraction drastically increases, temporarily boosting depletion rates. Thus, as most irrigation occurs in the drier parts of the world, where climate change is projected to lead to less rainfall, it can be expected that groundwater use will further increase, as surface water resources run out and irrigation water demand increases due to high temperatures. It is therefore likely that climate change will become an equally important driver for groundwater depletion as socio-economic change (see a recent review on groundwater and climate by Richard Taylor et al. in Nature Climate Change).

As I stated above, groundwater depletion occurs at several hotspots, so it seems like a regional problem. However, it has global dimensions for two reasons. First, if I buy a melon in my local Dutch supermarket it is likely to be imported from the south-east of Spain and will have been partly grown on non-renewable groundwater. Other examples of things I use or consume that may contain a non-renewable groundwater footprint (see the paper of Tom Gleeson et al in Nature 2012) are my jeans made from cotton or the meat I eat that comes from cows or pigs raised on soya-based concentrate. Second, as we and several other groups have shown, groundwater depletion is contributing to sea-level rise. During the second half of the 20th century, impoundment of water behind dams had a negative impact on global sea levels. However, the positive contribution of groundwater depletion became larger than the negative contribution by dams during the 1990s and groundwater is now the main positive terrestrial non-frozen water source, contributing 0.4-0.8 mm/year to the current sea-level rise (based on various estimates) and expected to go up to 1 mm/year by 2050. Obviously, it is not going to lead to the catastrophic sea-level changes that may be expected from ice loss from Greenland or Antarctica, but the contribution from groundwater has to be taken into account in attribution studies explaining current sea-level rise.

Although we all learn that groundwater resources account for over 97% of the available non-frozen freshwater resources on earth, most of the studies on water availability and water scarcity from the past two decades have focused on surface water. This is partly explained by the fact that we know very little about our global groundwater resources. First, current estimates of groundwater abstraction vary between 700-1500 km3/year, return flows from 400-800 km3/year and depletion rates between 130-330 km3/year. Second, there is no reliable information on how to attribute irrigation water withdrawal to groundwater or surface water and we are still looking for a rational way to model this attribution. Third, and most importantly, we really have little clue of how much groundwater there is in the world and how much of these resources are recoverable. There are few estimates of the total amount of groundwater on earth (most well known is the one by Shiklimanov: 10.4 million km3), but some estimates differ by a factor of 5. We as yet do not have a good hydrogeological map of the world, and although there have been recent efforts to map groundwater globally (see the recent paper by Fan et al. in Science) and permeability (Gleeson et al. in Geophysical Research Letters), information of aquifer depth and the fresh-salt water boundary is lacking. Global hydrogeological mapping and groundwater assessments are therefore necessary to estimate the time horizon by which recoverable groundwater in regions with high depletion rates will have gone or have become unattainable to small farmers.

Groundwater depletion is a regional problem with global consequences, but it needs local to regional solutions. Several lines of measures should be tried together, involving water technology, food technology, water management, water governance, water finance and perhaps cultural change. Water technological solutions involve the improvement of irrigation efficiency (from sprinkler to drip or subsoil irrigation) and recover losses and recycling (capture and use return flows in combination with partial purification). Food technology mostly involves the introduction of cultivars with higher water productivity or more specific, genetically engineered rice cultivars that do not need to be flooded or can grow in salt water. Here, the cultural dimension cuts in as well, as a change in diet from rice to wheat and maize would save a lot of water and will open up the possibility for much more efficient irrigation techniques. Water governance plays an important part as well. For instance, subsidizing electricity or diesel for groundwater pumps, as is done in certain regions in the world in order to boost food production, not only leads to a large increase in small agricultural wells, even in areas with surface water close by, but, in the event of flooding, also causes the pumps to be active night and day, irrespective of actual irrigation water requirements, thereby leading to huge losses. Water management solutions will possibly be along the lines of conjunctive groundwater and surface water use and the development of artificial storage and recovery (ASR) and artificial recharge (AR). These measures are specifically promising in areas with seasonal rainfall, and experiences in the US with the High Plains aquifer have been positive. Finally, financial arrangements that promote water efficient crops and irrigation technology are necessary to make this transition, as well as investments by e.g. public-private partnerships to start experimenting and building ASR and AR infrastructure.

In this context, a controversial question that should be asked is whether groundwater should be viewed as a renewable resource everywhere at all times. There are regions in the world with enormous fossil groundwater reserves. The most well-known example is the Nubian Sand Stone Aquifer in Northern Africa that is estimated to contain some 150,000 km3 of groundwater. Should this water remain untouched just because it is non-renewable in our time? Or should the region exploit this precious fossil resource (as other countries use oil) to aid the development the region that is likely to be hit hard by climate change and explosive population growth? These difficult issues can only be discussed rationally if we have enough information about how much groundwater we have in the world and what is its renewal rate. Research by the hydro(geo)logical community should urgently tackle these questions.

Ultimately, a mix of solutions should be developed in cooperative projects between international and regional experts and local stakeholders. Exchanging experiences between regions will certainly aid these development. On March 20 I will visit the Fifth Regional Consultation Meeting on Groundwater Governance of the UNECE region in The Hague. I am curious to find out what regional initiatives, experiences and solutions are out there.