The study, led by Ayse Ozer together with Afshin Shahbazi and Halit Apaydin, integrates CMIP6-based climate projections under the RCP 4.5 and RCP 8.5 emission scenarios with spatial interpolation techniques and a rich mix of meteorological and land use datasets. The goal was to project streamflow dynamics all the way to 2099, a horizon that stretches far beyond the capabilities of most operational forecasting systems. What sets the work apart is not merely the ambition of the timeline but the sheer scale of the model optimization effort. The team tested approximately 40,000 alternative configurations, spanning hyperparameter adjustments, architectural modifications and training strategies, in a systematic search for the most reliable predictors of future water availability.

The two modeling approaches at the heart of the study represent distinct strands of deep learning. LSTM networks, first introduced by Hochreiter and Schmidhuber in 1997, are recurrent architectures designed to capture long-term temporal dependencies through gated memory cells that regulate what information is retained, updated or discarded at each time step. This makes them naturally suited to hydrological time series, where the flow in a river today reflects rainfall events, snowmelt and soil moisture conditions accumulated over weeks or months. CNNs, by contrast, rose to fame in image recognition, most famously in the ImageNet classification breakthrough by Krizhevsky and colleagues. In hydrological applications, convolutional layers act as feature extractors, sliding filters across input sequences to detect local patterns and multi-scale structure in meteorological variables, effectively treating the input windows as one-dimensional signals rather than raw numbers.

Both architectures were trained to perform multi-step-ahead forecasts, meaning the models do not simply predict tomorrow’s flow from today’s conditions. Instead, they project streamflow across extended future horizons, a task that compounds error accumulation and demands that the networks learn genuine process dynamics rather than short-term autocorrelation. To give the models the richest possible picture of each sub-basin, the researchers enriched their inputs with spatially interpolated climate data and land cover variability, using techniques refined for sparse station networks to translate point measurements into spatially coherent fields. This spatial enrichment proved to be a decisive ingredient: deep learning models fed with interpolated inputs and land-use information significantly outperformed simpler configurations, indicating that hydrological intelligence depends as much on the quality and diversity of the inputs as on the sophistication of the network itself.

The performance results are striking. CNN models generally outperformed their LSTM counterparts in both sub-basins, achieving Kling-Gupta Efficiency (KGE) values between 0.80 and 0.85 during the testing phase. The KGE metric, widely regarded in hydrology as a more informative benchmark than the coefficient of determination, decomposes model skill into correlation, bias and variability components; values above 0.8 are typically interpreted as strong predictive capability. That a convolutional architecture, not classically designed for sequential data, edged out the temporally specialized LSTM underscores a recurring lesson in applied machine learning: with enough properly engineered inputs and careful tuning, feature extraction can matter as much as sequence modeling. The exhaustive configuration search, touching tens of thousands of design permutations, likely played a substantial role in pushing both architectures toward their performance ceilings.

But the study’s most consequential findings concern the future rather than the past. Once the optimized models were coupled with CMIP6-based climate projections under the two RCP pathways, the picture they painted of the Sakarya Basin toward 2100 was grim. In the E12A033 sub-basin, streamflow is projected to decline by approximately 37 percent under the moderate RCP 4.5 scenario and 42 percent under the high-emission RCP 8.5 scenario, compared to the historical baseline. In the E12A053 sub-basin, the decline is far more severe: reductions of 59 percent under RCP 4.5 and 62 percent under RCP 8.5. The divergence between the two sub-basins is itself instructive, showing that climate impacts on water resources will be spatially uneven even within a single river system, and that basin-by-basin assessment, rather than broad regional averages, is essential for planning.

The implications ripple outward from the Sakarya Basin. Türkiye sits in the Mediterranean climate zone, which the Intergovernmental Panel on Climate Change has repeatedly identified as a hotspot where warming and drying will outpace global averages. Declining streamflow of the magnitude projected here would strain drinking water supplies, hydropower generation and the irrigation that underpins agriculture in the region. The research suggests that when water availability in a sub-basin could be nearly halved, or worse, adaptive and data-driven management strategies are not optional refinements but necessities. Accurate long-term forecasts of the kind this framework produces give water managers something they rarely possess: a quantitative, locally calibrated expectation of what is coming, with enough lead time to restructure allocations, invest in efficiency and plan reservoir operations.

Methodologically, the study contributes to a rapidly growing literature pitting data-driven models against process-based hydrological simulators. Traditional physically based models require extensive calibration data and detailed knowledge of catchment characteristics, whereas deep learning approaches learn the rainfall-runoff relationship directly from observations. Recent comparisons in catchments across the United States and China have shown that machine learning techniques can match or exceed process-driven models in simulating streamflow, particularly when trained on rich, well-curated datasets. The Ankara team’s work extends this comparison into the climate-projection domain, demonstrating that deep learning can serve not just as a nowcasting tool but as a vehicle for translating global climate model output into actionable local hydrology through the end of the century.

The researchers are candid about the practical dimensions of their work. The R and Python code underlying the study and the datasets are available on request, and the framework’s reliance on publicly available CMIP6 archives, reanalysis-grade meteorological data and satellite-derived land use products means the approach is portable to other data-scarce regions facing similar climate stress. The combination of spatial interpolation to compensate for sparse station coverage with deep architectures tolerant of heterogeneous inputs offers a template for hydrological forecasting in the many parts of the world where dense monitoring networks remain a luxury.

As global emissions trajectories continue to determine which corner of the uncertainty envelope humanity inhabits, studies of this kind convert abstract climate scenarios into concrete, river-by-river consequences. The message from the Sakarya Basin is unambiguous: even under a moderate emissions pathway, the water flowing through these Turkish rivers by century’s end could be a fraction of what it is today. Whether that projection becomes a planning document or a post-mortem will depend on how quickly water authorities, in Türkiye and across the climate-sensitive Mediterranean, act on the warnings now flowing out of the neural networks.

The Arctic Great Rivers Observatory represents one of the most extensive and enduring monitoring programs dedicated to large northern river systems. Established nearly two decades ago at the Marine Biological Laboratory (MBL) in Woods Hole, ArcticGRO focuses on six major Arctic rivers: the Lena, Yenisey, Ob’, Mackenzie, Yukon, and Kolyma. These rivers collectively transport enormous volumes of freshwater and dissolved materials from vast continental watersheds across Russia, Canada, and the United States, discharging directly into the Arctic Ocean. This unparalleled dataset serves as a barometer for understanding how terrestrial and climatic shifts cascade through fluvial systems and ultimately affect oceanic processes in a region that is warming at twice the global average.

Suzanne Tank, currently an associate professor at the University of Alberta and former postdoctoral researcher at MBL, has played a central role in the ArcticGRO initiative. The project, initially termed PARTNERS, was conceived in 2002 by pioneering scientists Bruce Peterson, R. Max Holmes, and James McClelland, all affiliated with the MBL Ecosystems Center. Since its inception, ArcticGRO has been a keystone component of the U.S. National Science Foundation’s Arctic Observing Network, securing continuous funding and fostering an international collaboration that bridges disciplinary and national boundaries.

The research recognized by the Frontiers Planet Prize elucidates complex biochemical shifts in river water composition, reflecting broader Arctic environmental transformations such as permafrost thaw, changing hydrological regimes, and alterations in terrestrial vegetation cover. The long-term chemical monitoring undertaken by ArcticGRO integrates multiple parameters, including concentrations of nutrients, organic carbon, and trace metals, crafting a detailed “fingerprint” that encapsulates extensive biogeochemical processes occurring across diverse spatial scales. These signatures not only chronicle ongoing changes but also inform predictions of future ecosystem trajectories under a warming climate.

A unique strength of ArcticGRO lies in its multinational scope and the continuity of data collection. Coordinated efforts by scientists from Russia, Canada, and the United States underscore the importance of cross-border scientific diplomacy in addressing global environmental challenges. By building and maintaining a consistent time series spanning over twenty years, ArcticGRO provides unparalleled temporal context for assessing how climate-driven disturbances in individual river basins propagate effects downstream and into the Arctic Ocean, thereby impacting ocean chemistry, biology, and circulation patterns.

The monitoring of large Arctic rivers is particularly consequential given their dominant role in modulating the chemistry of the Arctic Ocean, which in turn influences global oceanic circulation and the planetary climate system. Alterations in riverine inputs can affect the stratification and nutrient availability in coastal waters, with cascading effects on marine food webs and the global carbon cycle. The work undertaken by ArcticGRO has been pivotal in quantifying these interconnected processes, advancing our mechanistic understanding of Arctic environmental dynamics in an era of rapid change.

James McClelland, the current director of ArcticGRO, emphasizes that the Frontiers Planet Prize recognition validates the value of long-term, collaborative research efforts. "This accolade underscores the power of sustained international partnerships for solving the pressing global challenges posed by climate change," he stated. "Our team’s commitment to building and maintaining high-quality datasets continues to be fundamental for detecting and interpreting subtle yet consequential shifts in Arctic ecosystems."

Since its launch, ArcticGRO has amassed a substantial archive of water samples and ancillary data, enabling retrospective analyses that leverage emerging analytical technologies. These archived samples offer invaluable resources for future research endeavors seeking to decode past environmental states and better forecast the multifaceted responses of Arctic ecosystems to anthropogenic pressures. The combination of fieldwork rigor, laboratory innovation, and cross-disciplinary collaboration exemplifies the ideal framework for addressing complex Earth system science questions.

MBL Director Nipam Patel lauded the ArcticGRO initiative, highlighting the exceptional legacy of MBL Ecosystems Center scientists in Arctic research for over fifty years. He noted, "The recognition by the Frontiers Foundation honors not just ArcticGRO, but all the dedicated efforts by MBL researchers to pioneer science that advances our understanding of global environmental change in critically vulnerable regions."

The Frontiers Planet Prize itself stands as the world’s largest science competition focused on safeguarding planetary health by accelerating innovative, scalable scientific solutions. Its 2025 National Champions, selected by a jury of 100 eminent experts in sustainability and planetary health, represent the forefront of global research aimed at forging a sustainable future. The contest’s substantial prize awards are designed to empower these champions to transition breakthrough science into impactful, real-world applications.

The scientific community and broader public alike are increasingly aware that the Arctic serves as a harbinger for planetary change. Studies such as those propelled by ArcticGRO not only document the transformations occurring at high latitudes but also provide key data streams necessary for guiding policy and management interventions. In the context of unprecedented atmospheric warming, permafrost degradation, and changing precipitation patterns, the continuous monitoring of Arctic river chemistry offers an indispensable lens on the trajectory of Arctic climate feedbacks and their broader consequences.

Ultimately, the Arctic Great Rivers Observatory exemplifies a model of how long-term, collaborative, and interdisciplinary science can unravel the complexities of environmental systems under stress. The recognition by the Frontiers Planet Prize not only celebrates a scientific achievement but also underscores the urgency and promise of sustained global stewardship in the face of accelerating climate impacts.

Subject of Research: Arctic river chemistry and its relationship to large-scale environmental change in the Arctic region.

Article Title: Recent trends in the chemistry of major northern rivers signal widespread Arctic change

Web References:

• Arctic Great Rivers Observatory: https://arcticgreatrivers.org/
• Frontiers Planet Prize: https://www.frontiersplanetprize.org/
• Nature Geosciences Article: https://www.nature.com/articles/s41561-023-01247-7

Image Credits: Jim McClelland

Keywords: Arctic ecosystems, Rivers, Scientific collaboration, Biogeochemistry, Climate change, Ocean chemistry