Climate Change & Water Resources
Assessing the impact of climate change on water resources is fraught with difficulties. Although the general tendencies are relatively straightforward, the magnitude and direction of change are expected to vary strongly across spatial and temporal scales. A major issue is the discrepancy between the global scope of climate models, and the local nature of water supply systems. This difference in spatial scale often results in a low predictive capacity of impact models. However, climate change is only one of many pressures on water resources. Being able to predict future impacts of climate change surely is a bonus. But even if projections offer limited information, many pathways to improved sustainability exist.
If one buys asparagus in London, it is likely it was produced in Peru. The country has become the largest exporter of asparagus in the world, also exporting many other agricultural products, ranging from cotton to bananas. What is remarkable though, is that these products originate from a narrow strip of land along the coast, squeezed between the Pacific Ocean and the Andes. It is one of the driest places on earth with precipitation typically below 20mm per year, about 30 times less than London. The success of Peru’s agricultural boom is rooted in water flowing from the nearby mountains. The orographic effect captures air moisture from the Pacific, while spill-over rains from the Amazon basin provide an additional source of water. Precipitation is stored in lakes, wetlands, and to a lesser extent in glaciers, from where it is released slowly and steadily into rivers and irrigation canals.
These water resources, however, may prove to be less reliable than they seem. Precipitation patterns over the Andes are governed by several large-scale circulation patterns, including the notoriously complex El Niño phenomenon. Little is known about the drivers of El Niño, but it is not unlikely that small perturbations will have strong local impacts on precipitation. Furthermore, increasing temperatures may alter the carbon content of wetland soils, affecting their water storage and regulation capacity. The fate of tropical glaciers is less uncertain, and does not forebode well for the local water resources1.
Water, as much as energy,
lubricates global economic
activity and sustains our
lifestyles. And thirsty we are.
Water, as much as energy, lubricates global economic activity and sustains our lifestyles. And thirsty we are. Although the minimal amount of water to survive is around three to four litres per person per day, a modern society uses magnitudes more. The daily water consumption in the UK is around 150 litres per person per day. Including the so-called virtual water, which is used to produce all goods and services we consume, the total amount of water required for a western lifestyle quickly rises to about 3m3 per person per day. Managing water resources is about balancing this demand with a supply that is as affordable and sustainable as possible. Climate change may pose significant challenges to this.
An Uncertain Future
Indeed, the water cycle is a fundamental part of our climate. Incoming radiation evaporates water from the oceans and the land surface, forming clouds and determining energy fluxes. The exceptionally large latent heat of evaporation makes water vapour one of the main carriers of energy around the globe. It comes as no surprise, then, that any perturbation to the global climate will have profound impacts on the water cycle, and by extension on water resources2.
Global climate change is expected to affect water resources in two fundamental ways: a change in the precipitation that replenishes freshwater resources, and a change in temperature that will cause more water to be evaporated from open surfaces and transpired by the vegetation. Combined, these are typically referred to as evapotranspiration. It is not just a result of the available energy; warmer air is also capable of storing more water vapour. From a water resources perspective, evapotranspiration is considered a loss. The part of precipitation that does not evapotranspire, either runs off the surface to rivers or infiltrates to recharge groundwater systems. Both rivers and groundwater aquifers can be exploited for water supply. The water vapour will also eventually condense and generate precipitation, but this may happen at a large distance from its source. So we can expect precipitation will rise globally, and this is indeed projected by global climate models. But behind these global trends lay strong spatial and temporal patterns, which may change in unpredictable ways.
Any perturbation to
the global climate will
have profound impacts
on the water cycle, and
by extension on water
Water resources are very vulnerable to local changes in precipitation patterns. Water is heavy, and therefore difficult to transport over large distances. As a result, water is sourced as close as possible to where it is to be consumed, which is typically within or close to the watershed. Especially in geographically complex areas such as mountains, the transport of water from one river basin to another is costly, as it requires digging tunnels and canals or even pumping water upslope. Hence, any changes in the spatial patterns of precipitation will affect water resources.
Temporal patterns of precipitation are expected to change as well. The higher energy content of the atmosphere will lead to a larger variability and stronger extremes. For precipitation this means longer and more severe droughts, as well as more intense precipitation events3. Both are bad news for water resources. Longer droughts will require more storage capacity, while intense precipitation events are more likely to damage and pollute water reservoirs and infrastructure than to help in replenishing them. At the same time, natural water stores such as vegetation, soils, wetlands and glaciers, are prone to degradation either by the changing climate itself or by other human activities4.
In the importance of the local precipitation patterns lays a fundamental issue with predicting the impact of climate change on water resources. Global climate models do not represent them well. This has partially to do with the coarse resolution of current model implementations. The spatial discretization of these models is typically of the order of several 100km, which does not capture local precipitation processes. But the stochastic nature of precipitation also makes it harder to predict than large-scale circulation patterns.
Projections of climate
change impacts on
water resources tend
to be riddled with
There are several ways to deal with the coarse resolution of global climate models. Regional climate models can be implemented at higher resolution, but they require a lot of data and computing resources. Alternatively, statistical methods can be used to translate large-scale impacts to the local scale, but they rely strongly on the assumption that the relation between large scale and small-scale processes remains stationary. Taking these uncertainties into account, the range of climate projections tends to be very wide, up to the point that there is no agreement between different models on the direction of the change in precipitation. The situation is aggravated by the need for impact models that translate the climate projections into variables of direct relevance for water resources, such as streamflow and groundwater recharge. These models typically add significant uncertainty to the final model projections.
As a result, projections of climate change impacts on water resources tend to be riddled with uncertainties. The question then remains if and how such projections can inform adaptation strategies and policy decisions.
No Regret Policies
Water managers usually turn to scientists for updates on the latest certainties. In the case of climate change, it might be wiser to turn to scientists for updates on the range of uncertainties for impact questions. Indeed, although climate and impact models are continuously improved, it is unlikely that the uncertainties in model predictions will decrease significantly in the near future. For every known model deficiency that is addressed, several unknown unknowns are discovered that complicate matters further. Still, adaptation decisions need to be taken now. The question is then whether prediction efforts riddled with uncertainties are helpful to optimise future water management. Developments in climate adaptation and water management research present a number of different approaches to uncertainties and decision-making.
Adaptive water management starts from the acceptance of irreducible uncertainties about future (climatic) changes5. It moves away from a predict-and-control paradigm towards a more adaptive approach, with continuous learning and flexibility as key aims. In this sense, infrastructure investments with high sunk costs, irreversible decisions, or fixed management strategies prevent continuous learning and adjustment. A more effective way of dealing with unpredictability is to avoid control by creating the capacity to respond effectively to changing and unknown conditions, through developing strategies that are robust under the full range of possible future scenarios, through diversification of strategies or through strategies that can be flexibly applied when needed.
from the acceptance
Robust strategies can be complemented with a focus on the key vulnerabilities of the water system and the services it provides, rather than on the optimal strategy. A related approach is the development of no regret interventions, defined as strategies that yield benefits regardless of future trends in climate scenarios. Given that climate is only one of the many uncertain processes, no regret strategies will favour measures that are beneficial for these other domains as well.
This approach is particularly useful in a water resources management context, where the uncertainties of other processes affecting water supply and demand tend to be much lower. For instance, the Andes have a long history of intensive land-use, with cultivation and grazing on steep slopes and large-scale deforestation leading to soil degradation. These processes are well known to affect water resources negatively. Soil compaction favours surface runoff, decreasing the recharge of groundwater aquifers and accelerating the hydrological response. The result is a river regime with higher peaks and lower base flows, increasing the risk for both flooding and water scarcity. At the same time, the population in the arid coast of Peru is growing fast. Lima, the second biggest desert city in the world, grew with a rate of over 2% per year over the last decade. With a degrading water supply and a steadily increasing demand, water resources are likely to run into trouble in the near future. Climate change may affect the velocity and magnitude, but will not change the fundamental trend. Hence, no regret strategies such as protecting ecosystems that provide clear benefits to water supply, such as high altitudinal wetlands, and reducing distribution losses and water consumption are obvious pathways to an increased sustainability.
These challenges are of course not limited to Peru. Tropical mountain areas are especially vulnerable because of their complex topography and climate, but water resources in many other regions are threatened by overexploitation and contamination. Managing these resources in a dynamic society with various quickly changing external pressures is challenging, and climate change will certainly increase this complexity. Being able to predict the potential impact of climate change, even with uncertainty, may help to reduce the options and inform decision-making. But even when uncertainties are high, there is a lot of scope in understanding how mankind currently interacts with the water cycle and exploring options to optimise this interaction.
The author would like to thank Dr. Art Dewulf from the University of Wageningen, Netherlands for input in the section on No Regret Policies.
Dr. Wouter Buytaert is an expert on water resources and environmental change and a lecturer at the Hydrology Group in Department of Civil Engineering at Imperial College London
 Bates B. C. et al. (2008) Climate Change and Water. Technical Paper of the Intergovernmental Panel on Climate Change. IPCC Secretariat.
 Pall P. et al. (2011) Anthropogenic Greenhouse Gas Contribution to Flood Risk in England and Wales in Autumn 2000. Nature. 470: 382-5.
 Buytaert W., Cuesta-Camacho F. & Tobon C. (2011) Potential Impacts of Climate Change on the Environmental Services of Humid Tropical Alpine Regions. Global Ecology and Biogeography. 20: 19-33.
 Pahl-Wostl C. (2007) Transitions Towards Adaptive Management of Water Facing Climate and Global Change. Water Resources Management. 21: 49-62.