Models factor out direct human influence

Climate change alters river flow

Water volumes in rivers worldwide have changed significantly in recent decades. An international research team including Goethe University has now been able to corroborate that climate change plays a critical role. Project coordinator was ETH Zurich.

Climate change influences the Earth’s water balance: Depending on region and season, it can lead to more floods or droughts and also have an impact on river flow. River flow is an important indicator of the water resources available to humans and the environment. How much water is regionally available also depends on additional factors, such as direct interventions in the water balance or land use: If, for example, water is channelled off for irrigation, if land use changes as a result, for instance, of deforestation or afforestation, or if dams are built, this also changes river flow.
How greatly river flow has changed in various regions of the world during the past decades has not yet been studied at global level on the basis of concrete measurement data. Equally, the question of whether changes visible around the globe can be traced back to climate change or to direct human influence has not yet been answered. An international research team led by researchers from ETH Zurich has now succeeded in disaggregating the influence of these factors. To do this, they analysed data from 7,250 gauging stations worldwide.
The study, which has now appeared in the journal Science, corroborates that the water volumes carried by rivers changed dramatically between 1971 and 2010. Complex patterns can be found: Some regions have become drier, such as the Mediterranean, southern Africa or the north-east of Brazil. Elsewhere, by contrast, water volumes increased, for example in Scandinavia.

Searching for the causes

The researchers investigated how these changes came about with the help of computer simulations, which they conducted in the framework of the international climate research network ISIMIP with the objective of studying the possible impacts of climate change. Overall, they used nine global hydrological models, into which they fed climate data from the period under study (1971 to 2010).
Geographer Dr Hannes Müller Schmied from Goethe University and the Senckenberg Biodiversity and Climate Research Centre was in charge of one of the models. “Model simulations are very important for interpreting measurement data and calculating various scenarios,” he explains, “because we can switch the influence of climate change and direct human influence on and off, so to speak, and compare the results with the measurement data.” The results of the model simulations tallied well with the analysis of the measurement data from actual rivers.
“This means that climatic conditions can explain the observed trends,” says Lukas Gudmundsson, climate researcher at ETH Zurich and first author of the study. In a second run, the researchers additionally included direct human-induced changes in their simulations in order to study the influence of these factors. This did not, however, alter the result. Changes in water and land use are thus evidently not the cause of changes in rivers worldwide. Water management and land use can indeed lead to major fluctuations in river flow at local level. “But what interested us wasn’t local trends but global ones that become visible over longer periods of time,” says Gudmundsson. That is why the researchers did not look at the data from individual gauging stations in isolation, but instead aggregated them in larger, subcontinental regions for their analysis. This made it possible to recognise the influence of climate change in the data.

Influence of greenhouse gases

The researchers were able to confirm the role of climate change using what is known as the attribution method: They compared their measurement data with simulations driven by output of climate models, which were calculated once with human-induced greenhouse gases and once without. In the first case, the simulation concurred with the actual data, but in the second case it did not. Without climate change, the observed changes would therefore probably not have occurred.
The study is the first to corroborate by means of measurement data that climate change has a visible influence worldwide on flowing water. “This was only possible thanks to the good collaboration among the researchers and institutions from twelve different countries involved in the project,” stresses Gudmundsson. The data collected from the 7,250 gauging stations worldwide were also a joint effort: The researchers collated them together with Australian partners in a previous study. They constitute the largest global data set on river flow available today.
“Thanks to the models, we can now calculate reliable scenarios of how major rivers will change in future under the influence of climate change,” says Hannes Müller Schmied. Such projections will provide an important basis for planning and thus help affected regions to safeguard water supply and adapt to climate change.

Publikation: Gudmundsson, L. et al. (2021): Globally observed trends in mean and extreme river flow attributed to climate change. Science