In a groundbreaking development, researchers have reconstructed long-term global terrestrial water storage anomalies, filling critical data gaps that have long hindered comprehensive hydrological analysis. This innovative work, led by N. Mandal from the Department of Civil Engineering at the Indian Institute of Technology (Indian School of Mines) in Dhanbad, India, promises to revolutionize how we understand and manage water resources, with significant implications for the energy sector.
The study, published in Earth System Science Data, addresses a persistent challenge in hydrology: the limited data duration from the Gravity Recovery and Climate Experiment (GRACE) and its follow-on mission (GRACE-FO). These satellites have provided invaluable insights into terrestrial water storage (TWS), but their relatively short operational periods have left significant gaps in our understanding of long-term water storage variations.
Mandal and his team have developed a novel approach to reconstruct TWS anomalies (TWSAs) from January 1960 to December 2022. By leveraging a combination of observed data, satellite information, and land-surface model outputs, they have created a complete dataset for the pre-GRACE era. The key to their success lies in a sophisticated machine-learning framework that identifies optimal predictors and simulates TWSAs with unprecedented accuracy.
“The integration of climate indices, such as the Oceanic Niño Index and Dipole Mode Index, along with land surface model outputs, has significantly enhanced our ability to predict TWSAs,” Mandal explained. “This approach not only fills the data gaps but also provides a more comprehensive understanding of hydrological extremes and climate change impacts.”
The researchers evaluated various machine-learning algorithms, including convolutional neural networks, support vector regression, extra trees regressor, and stacking ensemble regression models. The extra trees regressor emerged as the most effective algorithm for most grid cells globally, particularly in major river basins like the Ganga–Brahmaputra–Meghna, Godavari, Krishna, Limpopo, and Nile.
The implications for the energy sector are profound. Accurate prediction of TWSAs is crucial for managing water resources, which are essential for hydropower generation, cooling systems in thermal power plants, and maintaining water levels in reservoirs. “Understanding long-term water storage variations can help energy companies plan more effectively, mitigate risks associated with hydrological extremes, and optimize their operations,” Mandal noted.
The study’s findings are particularly noteworthy in river basins such as the Godavari, Krishna, Danube, and Amazon, where the simulated TWSAs outperformed existing models. The median correlation coefficient, Nash–Sutcliffe efficiency, and root mean square error values in the Godavari Basin, for instance, demonstrate the robustness of the new approach.
The researchers also assessed the uncertainty of their reconstructed TWSAs, finding smaller standard error magnitudes in arid regions compared to other areas. This insight can help in better water management strategies in arid zones, which are often critical for energy production.
The gridded dataset, featuring a spatial resolution of 0.50° × 0.50° and offering global coverage, is now available for public use. This dataset, published in Earth System Science Data, which translates to Earth System Science Data, represents a significant leap forward in hydrological science and has the potential to shape future developments in water resource management and climate change research.
As the energy sector continues to grapple with the impacts of climate change and increasing demand for water, this research provides a valuable tool for more informed decision-making. By bridging the data gaps and enhancing our predictive capabilities, Mandal’s work paves the way for a more sustainable and resilient future.