Irrigation’s climate effects and the water sustainability link

15 April 2011



There is evidence that expected warming trends from increased greenhouse gases have been locally ‘masked’ by irrigation-induced cooling. Two scientists at the NASA Goddard Institute for Space Studies, Michael Puma and Benjamin Cook, recently published a study entitled ‘Effects of irrigation on global climate during the 20th century’ in the Journal of Geophysical Research – Atmospheres. Here they explain their findings on how the rapid expansion of irrigation during the 20th century has affected the climate, as well as the uncertainty surrounding the future climatic effects of irrigation


While it is well known that the rapid expansion of irrigation during the twentieth century has significantly altered the Earth’s terrestrial hydrologic cycle, only recently has attention been focused on the connections between irrigation and climate. These irrigation-climate connections include possible changes in regional temperature and rainfall. It is therefore important to understand how 20th century changes in irrigation amount, extent, and location may have influenced climate, especially in the heavily irrigated regions of the world. Additionally, the insight gained from analysis of 20th century connections between irrigation and climate are potentially valuable for predicting future irrigation-climate interactions.

In some areas, recent studies have found that irrigation-induced cooling may be offsetting the warming that would be expected from increases in anthropogenic greenhouse gas (GHG). For example, a modelling study for California’s central valley recently concluded that the valley’s temperature trends were impacted by irrigation intensification, and that these trends were masking the expected warming from GHG. In light of this, it is important to understand how much warming may have been masked by irrigation, and whether this masking effect will be maintained, increased, or diminished in the future.

In our study, we explore the irrigation-climate connection using a global climate model (GCM) and 20th century irrigation data from a new reconstruction of global hydrography (see Wisser et al. [2010]). We modify the GCM to represent irrigation dynamics in a simplified way and assess – for the first time – the potential impacts of irrigation on climate over the course of the 20th century.

How can irrigation cool a region’s climate?

Irrigation involves the transfer of water from surface water bodies (lakes, rivers, and reservoirs) and groundwater aquifers to the land surface. This human modification of the water cycle can impact the exchanges of moisture and heat between the land surface and atmosphere – two primary ways by which the land surface may affect climate. Irrigation directly modifies these moisture and energy exchanges when it increases either direct soil evaporation or plant transpiration (collectively termed evapotranspiration).

Increases in evapotranspiration due to irrigation require additional energy, because solar energy is the force driving the conversion of irrigation water from liquid to vapour. When irrigation increases evapotranspiration, we have a shift in the surface energy balance; more energy is used for evapotranspiration, and less energy is used to heat the soil and near-surface air. We therefore would expect cooler temperatures near the surface, which is a response that has been documented extensively in observational and modelling studies. It is important to note, however, that irrigation only increases evapotranspiration when the amount of water in the soil limits the evapotranspiration rate (rather than atmospheric demand).

So why might a farmer irrigate if the soil contains adequate water? Rice crops – prevalent especially in Asia – require standing water. If these rice crops are grown in areas that are typically moist without irrigation, then evapotranspiration will not increase despite the addition of water to the land surface. Hence, we will also not see a direct irrigation-induced cooling.

The indirect changes in climate due to irrigation are also important. When irrigation directly increases evapotranspiration, more water vapour will be in the atmosphere. Additional atmospheric water vapour may produce changes in cloud cover, precipitation, and net radiation at the land surface. These changes have the potential to influence climate over relatively large regions. However, these indirect changes are poorly understood at present and more research is needed to understand these effects.

Global estimates

Irrigation expanded over the course of the 20th century, even more rapidly since the 1950s and 1960s. The global area equipped for irrigation was approximately 53M ha in 1901, increasing to 285M ha by 2002. While current irrigation may cause significant climatic changes at the regional scale, it is not known when and where these effects became significant during the 20th century. It also remains unclear how important irrigation has been relative to other changes in land use that affected a similar land area (eg urbanisation) during the last century.

A major challenge is to estimate irrigation globally, because observations of water use for irrigation are typically very limited. We used an irrigation dataset that combined observations with a water balance model to produce global estimates of irrigation for the 20th century. The first step in the development of this dataset was to identify ‘areas equipped for irrigation’ at the end of the 20th century. These areas are defined as agricultural lands that have built-in irrigation structures, although the structures might not always be in use. Next, a time series of irrigated areas was created extending back to 1901 by rescaling the year 2000 values with time series of irrigated area at the country level. In many countries prior to 1950, linear extrapolation was used, because data were unavailable.

Estimates of monthly irrigation rates were obtained using a soil-moisture balance for areas identified as ‘equipped for irrigation’. Irrigation water was applied to refill the soil to its holding capacity, whenever soil moisture dropped below a crop dependent threshold. For rice paddies, a 50mm layer of water is maintained under the assumption that water percolates at a constant, soil texture-dependent rate. Finally, we need to approximate the fraction of water used by crops relative to the amount withdrawn from irrigation sources. An irrigation efficiency factor, which accounts for water lost during distribution and field scale application, is estimated for different regions depending on typical irrigation practices. This factor allows us to compute the total amount of water withdrawn for irrigation, referred to as ‘gross irrigation’.

The average irrigation rates for June, July, and August for the early and late 20th century are presented in Figure 1. For the early part of the century, the largest irrigation amounts are in Asia (especially Pakistan and India), with secondary peaks in irrigation over central North America, Southern Europe, and Western and Southeastern Asia. By the end of the century, irrigation increases in these areas and becomes dramatically more extensive, especially in Northern Europe, Central Asia, and Western Asia.

Climate model

Our primary tool to explore the 20th century effects of irrigation on climate was the NASA Goddard Institute for Space Studies (GISS) GCM. The GISS GCM is a state-of-the-art atmospheric climate model, which we modified to include irrigation representation. The model modifications involved two main changes. First, irrigation water is withdrawn from the model’s lakes and rivers if available. If no surface water is available, we add water to the system, assuming that the water comes from a ‘fossil’ groundwater aquifer. The irrigation water is then added as a flux of water at the top of the vegetated soil (rainfall is treated in a similar fashion).

We used the GISS GCM to simulate climate (with observed sea surface temperatures) for the years 1902 to 2000. We conducted two sets of simulations, control runs without irrigation and runs with irrigation. Each set contains five runs (starting from unique initial conditions) to account for the inherit variability of the climate system. We then analysed changes to climate between the average climate of the irrigation simulation set (IRRIG) and the control simulation set (CTRL).

Impacts on 20th century climate

We found that irrigation-related climatic impacts generally increased over the course of the 20th century, especially after about 1950 when irrigation water withdrawals began to accelerate. As an example, temperature impacts were mostly co-located with irrigation regions (Figure 2), with cooling in many irrigated areas. At the same time, portions of China do not cool despite the relatively significant irrigation rates, because soil water is not limiting evapotranspiration in our control runs. Indirect effects of irrigation were also predicted – both cooling and warming in non-irrigated areas – as well as increases in down-wind rainfall and shifts in tropical rain regimes. While these indirect results are not robust and are poorly understood, they do appear to be important for understanding irrigation-climate connections.

We also found significant affects on regional temperature trends over the 20th century. In Figure 3, the regional effects of irrigation over the century are shown for Western North America and India. We find cooling in both regions, although this cooling is more substantial for India, likely due to the higher rates of irrigation. These findings support the possibility that irrigation may have masked the global warming signal in certain areas. As for rainfall, it increases slightly in Western North American, whereas India experiences overall rainfall decreases. These decreases are due to an indirect influence of irrigation. Specifically, the model predicts a weaker monsoon, because the temperature contract between the land and sea has been reduced.

Human hydrologic modifications

Our study showed that 20th century irrigation led to significant regional shifts in climate and also demonstrated the need for a greater understanding of irrigation-climate interactions. Likewise, other human modifications may also have important climate impacts. The next most important human hydrologic modification (besides irrigation) is the construction of dams. Dams, of course, all lead to the impoundment of freshwater runoff (ie reservoirs), which alters surface-water evaporation, apparent ageing of water on river systems, river-sediment retention, and nutrient cycling. All of these changes could be potentially important in terms of their effects on climate, such that future efforts should continue to explore how the human-modified hydrologic cycle affects the climate system. A better understanding of how water management decisions impact climate are ultimately important to inform issues related to water sustainability.

Dr Michael J Puma is a hydrologist studying the interactions of the hydrologic cycle with climate, terrestrial ecosystems, and society. He is currently an associate research scientist with the Columbia University Center for Climate Systems Research, which is associated with the NASA Goddard Institute for Space Studies Email: [email protected], and his research activities are described further at http://www.columbia.edu/~mjp38/.

Dr Benjamin I Cook is a research scientist at the NASA Goddard Institute for Space Studies interested in interactions between the land surface, terrestrial ecosystems, and the climate system. Email: [email protected], and his research activities are described further at http://www.ldeo.columbia.edu/~bc9z/.


Figure 3 Figure 3
Figure 1 Figure 1
Figure 2 Figure 2


Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.