top of page
Search

Increased moist-heat exposure caused by irrigation expansion

By: Yi Yao, Agnès Ducharne, Benjamin I. Cook, Steven J. De Hertog, Kjetil Schanke Aas, Pedro F. Arboleda-Obando, Jonathan Buzan, Jeanne Colin, Maya Costantini, Bertrand Decharme, David M. Lawrence, Peter Lawrence, L. Ruby Leung, Min-Hui Lo, Narayanappa Devaraju, William R. Wieder, Ren-Jie Wu, Tian Zhou, Jonas Jägermeyr, Sonali McDermid, Yadu Pokhrel, Maxwell Elling, Naota Hanasaki, Paul Muñoz, Larissa S. Nazarenko, Kedar Otta, Yusuke Satoh, Tokuta Yokohata, Lei Jin, Xuhui Wang, Vimal Mishra, Subimal Ghosh & Wim Thiery


Irrigation plays an important role in securing global food supply, and has experienced rapid expansion due to the increasing population and life quality (Figure 1). This expansion makes irrigation a climatic-effective land management, substantially affecting local climate patterns. Most studies revealed the cooling benefits brought by irrigation, ignoring the potential negative impacts of the increased air humidity.


Figure 1. Global and regional area equipped for irrigation (AEI) from the year 1900 to 2005 (Siebert et al., 2015).
Figure 1. Global and regional area equipped for irrigation (AEI) from the year 1900 to 2005 (Siebert et al., 2015).

Evaporation is an important way for human bodies to lose heat (Figure 2), and many moist-heat metrics have been developed to account for the impacts of air humidity on human comfort. Wet-bulb temperature is one of them which indicates the temperature when the maximum evaporative cooling is approached. In this study, through the joint efforts of multiple Earth System Modelling groups, we explored the impacts of irrigation expansion during the 20th century on irrigation water withdrawal and dry/wet-bulb extremes at the global scale (dry-bulb temperature is the temperature when no evaporative cooling is considered).


Figure 2. Energy balance in the human body (Buzan et al., 2020).
Figure 2. Energy balance in the human body (Buzan et al., 2020).

Results show that irrigation extent experienced rapid extension, from 0.5 million km2 to around 3 million km2 globally, which drives the substantial increase in irrigation water withdrawal. Increased irrigation water applied to cropland causes evaporative cooling, which could reduce the frequency of dry-heat extreme events by more than 4 times in some regions. However, alongside the evaporative cooling, air humidity is then also increased by irrigation, which may increase the frequency of moist-heat extreme events by >4 times. Averaged over the most intensely irrigated grid cells, the annual hours of dry-bulb (T2m) extremes are substantially reduced by irrigation expansion, but those of we-bulb (Tw) extremes are slightly enhanced (Figure 3). This study highlights the over-optimism of irrigation’s beneficial cooling in previous studies and calls for the attention to local people’s risk of exposure to moist-heat extremes.


Figure 3. a–c Time series of the annual hours over the grid cells with ≥ 40% of irrigated fraction increase (in the year 2014 compared to the year 1901) of 2-m air temperature (T2m: a), HUMIDEX (HU: b), and wet-bulb temperature (Tw: c) warm extremes. The warm extremes are defined as the period when T2m, HU, and Tw exceed their 99.9th percentile values of the first 30 years (1901–1930) in the simulations without irrigation expansion (1901irr). Lines indicate the median value among six models and ranges indicate the middle four of six models. Curves were smoothed using Savitzky-Golay filtering (order = 2, window = 15)72. The range of the y-axes are different for three sub-plots.
Figure 3. a–c Time series of the annual hours over the grid cells with ≥ 40% of irrigated fraction increase (in the year 2014 compared to the year 1901) of 2-m air temperature (T2m: a), HUMIDEX (HU: b), and wet-bulb temperature (Tw: c) warm extremes. The warm extremes are defined as the period when T2m, HU, and Tw exceed their 99.9th percentile values of the first 30 years (1901–1930) in the simulations without irrigation expansion (1901irr). Lines indicate the median value among six models and ranges indicate the middle four of six models. Curves were smoothed using Savitzky-Golay filtering (order = 2, window = 15)72. The range of the y-axes are different for three sub-plots.
"The post-processing of Earth System Model outputs was conducted on VSC supercomputer Hortense, which includes data transfer, analysis, and storage"

The post-processing of Earth System Model outputs was conducted on VSC supercomputer Hortense, which includes data transfer, analysis, and storage. VSC provided the data transfer option with GLOBUS, which helped us connect to other supercomputers easily. Then, the VSC team developed a platform on which we can easily use a Jupyter notebook to analyze the data on the system. Finally, iRods allows us to store all our data sets for a long-term period.



 

Siebert, S., Kummu, M., Porkka, M., Döll, P., Ramankutty, N., & Scanlon, B. R. (2015). A global data set of the extent of irrigated land from 1900 to 2005. Hydrology and Earth System Sciences, 19(3), 1521-1545.

Buzan, J. R., & Huber, M. (2020). Moist heat stress on a hotter Earth. Annual Review of Earth and Planetary Sciences, 48(1), 623-655.


Read the full article in Nature Communications here

bottom of page