Comparison of normalized storage (NS) variations of global pre- and post-1999 reservoirs: a Locations of the global reservoirs (with pre-1999 reservoirs in red and post-1999 reservoirs in cyan). b Comparison of the NS values of global pre- and post-1999 reservoirs (excluding regulated natural lakes), along with the accumulative storage capacity of the post-1999 reservoirs. Source: Study report.
Over the past two decades, global reservoirs have become increasingly empty despite an overall increase in total storage capacity due to the construction of new reservoirs, finds a new research. The study findings have been published in Nature Communications (Yao Li, Gang Zhao, George H. Allen, Huilin Gao. Diminishing storage returns of reservoir construction. Nature Communications, 2023; 14 (1) DOI: 10.1038/s41467-023-38843-5).
The decline in reservoir storage is particularly prominent in the global south, including South Asia, Africa and South America. Despite efforts to construct new reservoirs, the data shows that they fall short of expected filling levels.
The most significant decline is in South America and Africa, where growing populations contribute to an escalated water demand.
In contrast, reservoirs in the global north, including regions in North America and Europe, are experiencing an upward trend in reaching their maximum capacity. Reservoirs in high-latitude regions like the Great Lakes and Siberia exhibit comparatively higher storage capacities, primarily attributed to their lower population densities and lesser impacts from human activities.
The scientists doing the research used a new approach with satellite data to estimate the storage variations of 7,245 global reservoirs from 1999 to 2018.
Led by Dr. Huilin Gao, associate professor in the Zachry Department of Civil and Environmental Engineering at Texas A&M University, researchers used a new approach with satellite data to estimate the storage variations of 7,245 global reservoirs from 1999 to 2018.
Overall, global reservoir storage increased at an annual rate of 28 cubic kilometers, attributed to the construction of new reservoirs. However, despite these efforts, the data reveals that the rate of reservoir filling is lower than anticipated.
“As the global population continues to grow in the 21st century, surface water reservoirs are increasingly being relied on to meet rising demands in the context of a changing climate,” Gao said. “However, the amount of water available in reservoirs and its trends have not been well quantified at the global scale.”
The researchers developed the Global Reservoir Storage dataset, freely available online to benefit decision-makers and the wider science community. It represents a significant advancement in tracking global reservoir storage conditions.
Given the projected decline in water runoff and the rising water demand, the observed trend of diminishing storage returns from reservoir construction is expected to continue, potentially impacting water supplies with significant implications. These findings indicate that addressing future water demands cannot rely solely on constructing new reservoirs, emphasizing the need for novel management strategies.
“Through this research, we share a new perspective for reevaluating the socio-economic benefits of new reservoir construction and the tension between growing water demand and lessening water availability in developing countries,” said Dr. Yao Li, a Texas A&M former postdoctoral researcher who is currently a professor at the School of Geographical Sciences at Southwest University.
The analysis did not consider the sedimentation process, and therefore the overall storage decline presented in this study is conservative.
Other contributors to this research are Dr. Gang Zhao, a former postdoctoral fellow at the Carnegie Institute for Science in Stanford who is now a researcher at the Institute of Geographic Sciences and Natural Resources Research, and Dr. George H. Allen, assistant professor of Hydrology and Remote Sensing at Virginia Polytechnic and State University in Blacksburg, Virginia. Both Li and Zhao are former Texas A&M students who worked in Dr. Gao’s research group, Gao Hydrology Research Group.
The study report said:
Surface water reservoirs are increasingly being relied upon to meet rising demands in the context of growing population and changing climate. However, the amount of water available in reservoirs (and the corresponding trends) have not been well quantified at the global scale. Here we use satellite observations to estimate the storage variations of 7,245 global reservoirs from 1999 to 2018. Total global reservoir storage has increased at a rate of 27.82 ± 0.08 km3/yr, which is mainly attributed to the construction of new dams. However, the normalized reservoir storage (NS) — the ratio of the actual storage to the storage capacity — has declined by 0.82 ± 0.01%. The decline of NS values is especially pronounced in the global south, while the global north mainly exhibits an NS increase. With predicted decreasing runoff and increasing water demand, these observed diminishing storage returns of reservoir construction will likely persist into the future.
It said:
The 20th Century witnessed a massive dam construction boom, first starting in North America and then spreading to the rest of the inhabited world. Behind these dams, sprawling reservoirs have fundamentally enhanced our ability to manage Earth’s freshwater resources, but have also imposed adverse environmental and social effects. After a decline in growth during the 1990s, hundreds of large new dams have been added in Asia, Africa, and South America. In addition, over 3700 hydropower dams (each over 1 MW in capacity) are being planned or are under construction as of 2014, most of which are in developing countries. With water scarcity intensified by both climate change and increasing water demand, reservoir water availability is essential for sustainable development. Yet dam construction and reservoir operations are rarely coordinated amongst countries despite nearly half of all land being covered by international river basins. To best inform future decision-making related to global surface water management, the storage conditions of reservoir impoundments — particularly those newly constructed — should be carefully evaluated.
However, knowledge about the long-term variation of reservoir storage is very limited at a global scale. In situ measurements of reservoir storage are often not shared, especially across international river basins. Land-surface and hydrologic models produce highly uncertain storage estimates, largely due to the lack of reservoir operations/management information. By relating reservoir surface area and elevation values, satellite remote sensing provides a viable alternative for monitoring storage. Although recent studies have quantified long-term surface area time series values and seasonal elevation variations of reservoirs globally, reliable storage estimations have only focused on reservoirs built before 1999 (hereafter referred to as “pre-1999 reservoirs”).
The scientists have introduced “normalized storage” (NS), a new term defined as the ratio of the actual storage from a group of reservoirs to the storage capacity of these same reservoirs. The NS offers a unique flexibility for quantifying and comparing the storage returns of impoundment from different groups of reservoirs across global, continental, and basin scales. This makes regions with different storage capacities comparable, and it is not affected by the increased storage from new reservoirs. The use of NS allows us to: (1) split the pre- and post-1999 reservoirs and directly evaluate the behaviors of newly constructed reservoirs; (2) compare the NS trends with the actual storage trends for new insights; and (3) assess reservoirs grouped by different functionalities (e.g., hydropower, irrigation).
While the newly constructed reservoirs have contributed to a steady increase in global storage capacity, their storage returns (in terms of NS) are found to be smaller than those built in the 20th century (over the period of 1999–2018). The NS values of the post-1999 reservoirs are significantly lower (and with larger seasonal variations) than the pre-1999 ones, at 60.59 ± 4.33% and 70.36 ± 1.36%, respectively. The NS values also depend on reservoir function. For instance, the NS values in reservoirs whose primarily function is hydropower are generally higher than those whose primarily function is irrigation or flood control. Therefore, the scientists further compared the NS values for pre-1999 and post-1999 reservoirs in terms of reservoir function, at both global and basin scales. Regardless of the function and/or spatial scale, all results lead to the same conclusion — that post-1999 reservoirs have lower NS levels, but larger seasonal variations, than pre-1999 reservoirs. It is worth noting that there are many social-economic benefits to building reservoirs (e.g., hydropower generation, flood reduction, water supply, and recreation), but here we are framing the returns only in terms of normalized water storage.
The study report said:
Basin-scale reservoir storage information is essential for managing local water resources and assessing changes to the hydrological cycle, yet this knowledge is lacking at the global scale — particularly for transboundary river basins. Thus, we evaluate the storage variations at the basin scale. The majority of the Earth’s basins experienced storage growth. Asian basins have the most storage growth, while basins in southern Africa have suffered from storage losses. The highest increase is found in the Yangtze River Basin (3.34 ± 0.01 km3/yr), which is characterized by the most intensive dam construction activities in Asia (e.g., Three Gorges Dam). The fastest storage decline occurred in the Colorado River Basin (−0.62 ± 0.003 km3/yr), due to the combined effects from an extended drought since 2000 and increasing water use. With regard to NS, decreasing trends predominate — especially over the Southern Hemisphere, including those areas in South America and southern Africa. The basins with increased NS are mainly in East Asia, Europe, and North America. However, fewer basins in Asia show significant increases in NS compared to storage — and the NS increases in European and North American basins are generally weak.
The reservoirs in high-latitude regions (e.g., the Great Lakes and Siberia) have relatively high NS values. This is attributable to the fact that these northern reservoirs are less affected by human activities due to low population density. On the other hand, the basins in South and Southeast Asia (e.g., the Indus and Yangtze basins) have low NS levels likely because of the high water demand driven by the large populations in these areas. Moreover, the annual mean coefficient of variation (CV) of the NS values indicates that the reservoirs in high-latitude areas — as well as those in Europe and North America — are relatively stable, with comparatively small dynamics. In contrast, the Asian reservoirs — with the exception of those in high-latitude regions — show large intra-annual variability. The large NS variability in the Amazon basin is attributable to the extensive damming of highly seasonal Amazonian rivers. Finally, the basin with the highest annual CV value — the Indus basin in India — has experienced substantial groundwater depletion, suggesting that surface water and groundwater levels have both been dominated by human activities in response to the region’s severe water shortage.
It said:
Reservoir NS is impacted by multiple factors, including upstream runoff, population density, and reservoir function — which directly affect reservoir storage values through inflow, demand, and operation. The global runoff, especially in tropical regions, suggests a considerable decrease during the last two decades. This suggests that declines in runoff may be the driver of the observed trends (of decreasing NS) in reservoirs with large storage variations. The decreasing trends in runoff are most significant in South America followed by Africa, which contribute to the reduction of NS values in these regions. These trends may exacerbate already stressed reservoirs where NS is significantly decreasing. In the last two decades, global population growth is outpacing increases in reservoir storage, causing a significant decrease in the per capita storage (−0.38 ± 0.04 m3/yr). If regionally declining runoff persists into the future, it will likely add considerable pressure on global reservoirs for meeting the anticipated rising demands on water resources. In particular, we expect that regions in Africa and South Asia will experience increasing water shortages under the predicted extensive decrease in runoff and increase in population. Special attention should be paid to reservoir water management strategies in these hotspots of water scarcity.
Global water stress under population growth
a Runoff (Q) trends from 2000 to 2018 at the basin scale, with the significant trends (p < 0.05) delineated in white and the non-significant trends (p > 0.05) delineated in black. b Population density trends from 2000 to 2020 at the basin scale, with the significant trends (p < 0.05) delineated in white and the non-significant trends (p > 0.05) delineated in black. c The world population, global reservoir storage, and per capita storage from 1999 to 2018. Shading illustrates the 95% confidence intervals for the best-fit linear trends, and ʻaʼ represents the trend value. The world population data were collected from the World Bank (https://data.worldbank.org/indicator/SP.POP.TOTL), and the annual total reservoir storage values were derived by averaging the monthly Global Reservoir Storage dataset. The per capita storage is the ratio of global reservoir storage and world population.
The scientists’ analysis reveals:
The global reservoir normalized storage (i.e., NS) has significantly declined in the 21st century despite an increase in total storage due to the construction of new reservoirs. The changes mainly occurred in South America, and Africa, where the world’s developing countries are located. However, these negative trends have been weakened at the global scale due to the significant increasing NS in the Global North — North America (2.61 ± 0.01%/20 yr) and Europe (1.50 ± 0.01%/20 yr). In particular, the NS values of the post-1999 reservoirs are much smaller than those of the pre-1999 ones. Asia, South America, and Africa are the primary “hotspots” where most future dam construction activities are planned, with Brazil, China, and the Democratic Republic of Congo taking the lead. However, future development of new reservoirs likely will not alleviate the water stress caused by increasing municipal and industrial water demand in south Asia (e.g., India) and southeast Asia (e.g., China). In South America and Africa, the reduced NS values are mainly associated with the large decreases in runoff trends. The results from this study highlight the challenges of resolving finite water resources through reservoir regulation — particularly in developing countries, where the storage returns from these impoundments are diminishing. These findings offer a new perspective for reevaluating the socio-economic benefits of new reservoir construction, and the tension between growing water demand and lessening water availability in developing countries.
The scientists collected in situ measurements for 277 reservoirs from the United States, Australia, and India to validate the storage results (a total of 101,041 pairs).
Note:
Area-Storage = A-V
Global Reservoir and Dam Database = GRanD
Small hydropower plants =SHPs
Global Reservoir Surface Area Dataset = GRSAD
Surface Water Occurrence = SWO
