The Grand Ethiopian Renaissance Dam (GERD) is a significant hydroelectric dam under construction on the Blue Nile, a major tributary of the Nile River. Its construction, coupled with the existing High Aswan Dam (HAD) in Egypt, presents unprecedented challenges and opportunities for water resource management in the Nile Basin. The Nile, a lifeline for millions, is characterized by high inter-annual variability and significant downstream consumptive use, primarily in Egypt and Sudan, despite 85% of its water originating in Ethiopia. The GERD, primarily intended for power generation, will substantially alter downstream flow patterns, raising concerns about water security and the need for an agreement on basin-wide management. The 1959 Agreement between Egypt and Sudan, which initiated the construction of the HAD, does not involve Ethiopia and other upstream countries, making negotiations on the GERD's operation complex. This study aims to analyze potential scenarios, quantify risks, and highlight the importance of cooperation among riparian nations for effective water resource management.

Existing literature on Nile hydrology highlights the river's high inter-annual variability and the impact of existing infrastructure. Several studies have examined the trade-offs between riparian objectives and the benefits of basin-wide cooperation concerning the GERD and HAD operations. Previous research offers valuable insights into potential risks and rewards of cooperation, informing multilateral agreements and strategies for addressing diverse interests within the basin. However, a comprehensive analysis encompassing the filling period, the new normal operational state, and the management of potential droughts is still lacking. This study builds upon existing work to provide a holistic risk assessment under diverse hydrological conditions.

This research employs the Eastern Nile River Ware Model (ENRM), a detailed analytical water resources systems model, to simulate the Nile system from 2020 to 2060. The ENRM, developed using the River Ware platform, simulates reservoir operations based on prioritized logical statements that reflect the objectives of each water management infrastructure. These objectives include meeting agricultural, municipal, and industrial water needs, power generation targets, sediment management, minimum flow requirements, and drought management policies. The model incorporates historical hydrological data, including naturalized inflow reconstructions, to represent a range of hydrological conditions. Three distinct eras are simulated: (1) the GERD filling period (Era 1), (2) the new normal after filling (Era 2), and (3) a severe multi-year drought (Era 3). Era 1 considers a filling policy based on a proposal by the National Independent Scientific Research Group (NISRG). Era 2 examines a typical 20-year sequence of flows, assuming a baseload hydropower production of 1600 MW from the GERD. Era 3 simulates the 1972-1987 drought sequence to assess drought response strategies and the subsequent recovery period. Key assumptions regarding water withdrawals, evaporation losses, and reservoir operation rules are detailed in the Supplementary Information. Sensitivity analysis was conducted on critical assumptions to ascertain the robustness of model results.

The simulation results reveal several key findings: **Era 1 (GERD Filling):** The HAD reservoir levels may fall to unprecedented lows during the filling of the GERD, even though the risk of water shortage in Egypt remains relatively low. This is largely due to the existing high storage levels in the HAD before the filling begins. **Era 2 (New Normal):** Post-filling, Ethiopia and Sudan experience significant benefits with no substantial negative effects on Egypt under average flow conditions. Ethiopia benefits from consistent hydropower generation, while Sudan experiences smoother flows, reduced flood losses, and increased irrigation opportunities. **Era 3 (Severe Drought):** A severe multi-year drought, such as the 1970s and 80s drought, would have significant negative consequences for all three countries. While the GERD initially moderates shortages in Egypt, both the GERD and HAD reservoirs become severely depleted. Re-filling the reservoirs after such a drought presents significant coordination challenges and potential for conflict due to differing priorities (hydropower generation vs. consumptive use). The refilling stage of such a drought creates a greater water deficit to Egypt than the filling of the GERD. This illustrates the inherent need for cooperative agreements to prevent potentially detrimental consequences. The study also highlights the potential for 'water panic'—a public perception of water insecurity that could exacerbate existing tensions and hinder effective drought management.

The findings demonstrate that the GERD's impact on the Nile system is complex and context-dependent. While the filling period may cause concerns due to the visual impact of decreasing HAD reservoir levels, it is unlikely to cause significant water shortages in Egypt under current management strategies. The new normal, however, underscores the need for cooperation, particularly in managing potential droughts. The risk of severe multi-year droughts, exacerbated by climate change, necessitates a proactive and coordinated approach. The model's results highlight the critical importance of building trust and fostering transparent communication among riparian nations to prevent misunderstandings and potential conflicts during both the filling phase and inevitable future drought periods.

This study emphasizes the need for cooperative transboundary water management in the Nile Basin. The simulations highlight the varying impacts of the GERD under different hydrological conditions and underscore the importance of proactive planning for managing the inevitable multi-year droughts. Cooperative agreements on filling strategies and drought management protocols are crucial for ensuring water security and preventing potential conflicts. Future research should focus on incorporating more sophisticated climate change projections into hydrological forecasting and explore further the behavioral economics aspects of water panic and its management.

The model's results rely on several key assumptions, including water withdrawal patterns, evaporation losses, and reservoir operation rules. Uncertainties in these assumptions could affect the precision of the predictions. The study primarily focuses on historical hydrological sequences and may not fully capture the potential impacts of future climate change, which could lead to more extreme hydrological events. Finally, the model does not explicitly incorporate political and social factors that could further influence water allocation decisions during periods of stress.