Variable renewable energy pathways in the Lower Mekong Basin under projected river flow extremes

Abstract

The Lower Mekong River Basin faces growing challenges in deploying variable renewable energy under intensifying hydroclimatic extremes. To evaluate these impacts, we develop an integrated modeling framework that links an energy expansion model with a synthetic streamflow generator. The results indicate that severe extended periods of low flows (10,000-year events) reduce hydropower output by 24.3 ± 8.7 terawatt-hour (−13.7% ± 4.9%), increasing wind capacity requirements by 10.6 ± 12.5 gigawatts (1.15% ± 1.36%) and raising system-wide costs by 8.7 ± 3.2 billion U.S. dollars. Conversely, extended periods of high flows of similar severity boost hydropower generation by 23.3 ± 9.2 terawatt-hours (+ 13.1% ± 5.2%), decreasing wind capacity needs by 9.3 ± 11.7 gigawatts (1.02% ± 1.27%) and lowering system costs by 7.8 ± 3.0 billion U.S. dollars. Prolonged drought conditions accelerate wind development to sustain decarbonization, while surplus flows from floods may delay investment—introducing long-term uncertainty for energy system resilience. These findings underscore the need for proactive, region-specific strategies to ensure resilient and sustainable energy transitions under intensifying hydroclimatic extremes.

Introduction

Hydropower has long been a reliable and cost-effective renewable energy source, contributing to economic development in regions with abundant river systems¹. In the face of growing climate variability and change, hydropower plays an increasingly critical role in supporting global clean energy transitions and national decarbonization goals^2,3,4. In addition to electricity generation, hydropower provides vital grid flexibility by helping to balance the intermittent output of variable renewable energy (VRE) sources such as solar and wind^4,5. Currently, hydropower accounts for 36% of the world’s installed renewable energy capacity and supplies more than 14% of global electricity demand⁶. Its strategic importance within energy systems—and within the broader water-energy nexus—is expected to continue well into the 21st century⁷.

However, the reliability of hydropower is increasingly undermined by hydroclimatic variability, which disrupts generation and complicates the integration of VRE⁸. Such disruptions often stem from extended periods of low flows (EPLFs) or extended periods of high flows (EPHFs), both of which jeopardize the stability of hydroelectric output⁹ and challenge efforts to meet low-emission energy targets¹⁰. In particular, extended low-flow events, typically related to drought, threaten hydropower reliability and regional energy security¹¹. In 2023, for instance, global hydroelectricity shortfalls caused by drought-induced low flows led to an estimated 170 million tonnes of additional CO₂ emissions, as power systems turned to fossil fuel-based generation to compensate¹².

While high river flows can boost hydroelectric generation, extreme flooding poses serious risks to infrastructure, potentially causing structural damage and disrupting energy production¹³. Hydropower operators face the complex task of balancing flood control with energy optimization: excessive caution may waste generation opportunities, while insufficient mitigation heightens the risk of costly damage and outages¹⁴. Herein, we are concerned with extended periods of high flows (EPHFs)—sustained increases in annual river discharge that can enhance hydroelectric output over longer timescales. Beyond immediate energy gains, EPHFs may influence the broader energy mix by reducing reliance on variable renewable energy (VRE) sources and altering deployment trajectories. These constrasting flow dynamics emphasize the need for integrated, sustainable solutions that align hydropower operations with long-term energy and environmental objectives. Our analysis focuses on the long-term impacts of climate shocks—including EPLFs and EPHFs—on VRE energy generation and development.

Existing studies highlight considerable uncertainty in the future availability and economic value of hydropower under a changing climate¹⁵. Hydrological projections using global hydrological and general circulation models offer insights into potential river flow changes across climate scenarios, but they remain limited by inherent uncertainties¹⁶. For instance, Liu et al.¹⁷ found that climate change will have uneven impacts on China’s hydropower potential, with some regions gaining water resources while others experience sharp declines. These possibilities underscore the growing risk of declining hydropower reliability in key regions, reinforcing the need for diversified energy portfolios, improved water resource management, and adaptive, sustainability-oriented infrastructure planning—particularly in regions vulnerable to climatic extremes^18,19. A key emerging issue is the extent to which extreme climate events will disrupt hydropower generation²⁰. However, model-based assessments are constrained by the complexity of climate systems and the biases inherent in individual models¹⁰.

Integrating climate extremes into energy planning is complicated by the tight interdependence of water and power systems within large reservoir networks, where hydropower operations can trigger non-linear, cascading impacts^21,22. These interlinkages within the water-energy nexus demand more comprehensive planning frameworks. Moreover, simplified reservoir models often underestimate the flexibility value of hydropower in balancing VRE, leading to inflated projections of capacity needs and system costs^2,4. Addressing these challenges calls for more integrated modeling approaches that incorporate reservoir dynamics, account for system-level sustainability, and capture the uncertainties posed by extreme climate conditions.

To investigate optimal energy expansion pathways under extreme climate shock scenarios, we couple a hydropower-centric energy expansion model—PREP-SHOT (Pathways for Renewable Energy Planning coupling Short-term Hydropower OperaTion)—with a synthetic streamflow generator (Fig. 1).

Fig. 1: Conceptual framework for assessing the impacts of climate-related extremes on energy decarbonization expansion pathways within the Lower Mekong River Basin (LMRB).

This framework consists of three main parts. For part 1 (that is data preparation), we first characterize the cascade topology of all reservoirs and collect the historical daily streamflow data of the basin outlet, and then reproduce key stream statistics, such as seasonal means and variances while preserving spatial cross-correlations across reservoirs through a sets of process including the Cholesky decomposition, bootstrap resampling and percentile selection (see the section on Synthetic streamflow generation for more details), and finally generate synthetic streamflow for four representative years (2020, 2030, 2040, and 2050) across all 57 reservoirs within the LMRB under seven different climate scenarios, each comprising 100 ensemble members. For part 2 (that is model solving), we drive the Pathways for Renewable Energy Planning coupling Short-term Hydropower OperaTion (PREP-SHOT) model using the synthetic streamflow data under different climate scenarios associated with other related parameters and subsequently obtain the energy expansion pathways under different climate scenarios. For part 3 (that is scenario analysis), we compare the different outputs derived from the PREP-SHOT model and further evaluate the climate extremes’ impacts on renewable energy development within the LMRB. EPLFs and EPHFs represent extended periods of low flows and extended periods of high flows, respectively.

We simulate daily inflow time series for 57 existing reservoirs in the Lower Mekong River Basin (LMRB; Fig. 2) for four time slices (2020, 2030, 2040, and 2050), under baseline conditions and during extended periods of altered river flow corresponding to 100-, 1,000-, and 10,000-year return periods (shown in Table 1). Each scenario is run 100 times to capture variability. PREP-SHOT is then used to optimize future power generation mixes and identify the necessary expansion of electricity transmission infrastructure within the basin. Our analysis centers on the influence of EPLFs and EPHFs on solar and wind power deployment. Finally, we quantify the monetary impacts of these climate extremes on long-term energy system planning in the LMRB.

Fig. 2: Interconnected water and power grids in the Lower Mekong River Basin (LMRB).

A Full spatial extent of the Mekong River Basin and existing reservoirs located in the LMRB. B Spatial representation of the existing power system infrastructure for each country included in the LMRB. Pie charts show respective technology shares with areas sized by country-level total generation capacities.

Table 1 Climate shock scenario design

The LMRB serves as a model system due to its transboundary nature, extensive hydropower infrastructure, and ongoing debates over renewable energy development pathways²³ (Fig. 2). According to the Mekong River Commission (MRC), the basin had over 13.3 GW of installed hydropower capacity in 2023, supplying approximately 25% of the region’s annual electricity generation⁶. The LMRB is also experiencing rapid growth in wind and solar capacity²⁴, driven by economic expansion and the push toward a low-carbon energy transition²⁵. Cross-border electricity trade—driven largely by hydropower exports from Laos to Thailand and Vietnam—illustrates the region’s growing energy integration, now increasingly complemented by expanding interest in solar and wind power, bilateral power purchase agreements, and proposed high-voltage direct current interconnections²⁶. However, mounting evidence suggests that climate extremes will become increasingly frequent in the coming decades, threatening regional energy security²⁷. Our approach offers an analysis framework applicable to other international river basins with substantial VRE potential.

Results

System-wide energy capacity needs and cost changes

Under the modelled baseline scenario, the model framework estimates the addition of 1697 GW to the LMRB as optimal non-hydro renewable energy capacity between 2020 and 2050 (Fig. 3A; Table 2). This trend broadly aligns with national energy mix projections, though regional differences persist. Solar and wind dominate the expansion, contributing 446 GW and 918 GW, respectively, while the remaining 333 GW comes from other sources, including coal, oil, gas, bioenergy, and storage technologies such as pumped hydro and lithium-ion batteries.

Fig. 3: Impact of different climate extremes on total energy capacity needs and system cost.

A Ridgeline plots showing the total energy capacity needs under different climate extremes. B Changes in total hydropower production and total energy capacity needs in the LMRB during the planning horizon across all extreme climate shock scenarios relative to baseline scenarios. 1σ represents one standard deviation of all estimates of 100 realizations for each climate shock scenario. C Variations in total energy capacity needs in the LMRB and the major components during the planning horizon, including changes in Vietnam and the other countries under 10,000-year EPLFs and EPHFs relative to baseline scenarios. D Same as (C) but showing the changes in total system costs (denoted as ‘Sum_x’) during the planning horizon, which are primarily made up of three components: changes in investment costs associated with total energy capacity needs (denoted as ‘Capacity’), changes in fuel costs (denoted as ‘Fuel’), and the other costs (denoted as ‘Others’), mainly including fixed operation and maintenance (O&M) costs and variable O&M costs.

Table 2 Ensemble mean values of new installed capacity (GW) within the Lower Mekong River Basin (LMRB) and individual countries under normal conditions over the entire planning period (2020 to 2050)

Vietnam leads the region in wind energy expansion, contributing 469 GW—over half of the LMRB’s total wind additions. Thailand and Myanmar also prioritize wind, adding 218 GW and 176 GW, respectively. Myanmar pursues a more diversified strategy, investing an additional 94 GW in solar and 119 GW in other sources. Laos focuses almost exclusively on solar, installing 85 GW with minimal reliance on other energy types. Cambodia adopts a balanced portfolio, adding 94 GW of solar, 55 GW of wind, and 59 GW from other renewable and storage sources.

In this regional energy transition, characterized by a decisive shift toward VRE to meet decarbonization goals, hydropower capacity has been held constant. This reflects both concerns over overdevelopment in countries like Vietnam and increasing awareness of hydropower’s environmental and social impacts on riverscapes and local communities²⁸.

Modelled extreme changes in river flows disproportionately affect hydropower generation potential (Supplementary Fig. 1), leading to downstream adjustments in installed energy capacity. For EPHFs, we assess only the effects of sustained elevated river discharge, excluding short-term flood impacts, regardless of the magnitude or damage produced. Over the 2020–2050 planning horizon, EPLFs result in modest but consistent increases in system-wide energy capacity requirements in the LMRB relative to baseline conditions (Fig. 3A). Under 100-year EPLF events (annual probability = 0.01), total capacity needs rise by 3.7 ± 9.3 GW—an increase of 0.22% ± 0.55% compared to the baseline total of 1,697 GW. These estimates reflect the mean outcomes across 100 realizations per scenario (see Methods). In contrast, 100-year EPHFs reduce capacity requirements by 3.3 ± 9.0 GW, or 0.19% ± 0.53% below the baseline, due to increased hydroelectric production resulting from higher flows (Fig. 3B; Supplementary Fig. 2).

The severity of climate-related extremes scales inversely with hydropower production and directly with total energy capacity requirements (Fig. 3B). As EPLFs intensify from 100-year to 10,000-year events, system-wide energy capacity needs are projected to more than double, reaching 8.0 ± 9.8 GW (Fig. 3B). Conversely, as EPHFs increase in severity over the same return periods, total capacity requirements decline further, with reductions approaching 5.2 ± 7.7 GW—nearly twice the reduction observed under 100-year EPHFs (Fig. 3B).

At the country level, the most consistent changes in capacity needs across the LMRB occur in Vietnam (Fig. 3C), which relies heavily on hydropower imports from Laos to meet its growing electricity demand (Supplementary Figs. 3 and 4). However, hydropower production in Laos is particularly vulnerable to climate extremes, in part due to its large reservoir capacity and associated operational challenges—posing substantial risks to Vietnam’s energy security (Fig. 2). Under 10,000-year EPLF scenarios, our simulations indicate that reductions in Laos’ hydropower output account for 68.0% of the total LMRB-wide decrease in hydropower generation (−24.3 ± 8.7 TWh), equivalent to a cumulative loss of 16.5 TWh over the 2020–2050 planning horizon (Supplementary Fig. 2). This shortfall in Laos propagates to Vietnam—its largest electricity trading partner, leaving it increasingly exposed to upstream climate risks (Supplementary Fig. 3B).

In addition to reduced electricity imports from Laos for the EPLF scenario, Vietnam faces further energy deficits due to a substantial decline in its own hydropower generation, which accounts for 25.0% of the total LMRB reduction—equivalent to 6.1 TWh (Supplementary Fig. 2). Offsetting these deficits requires substantial energy capacity expansion, resulting in a net increase of 8.0 ± 9.8 GW in total system-wide capacity relative to baseline conditions (Fig. 3C). Most of this growth is driven by the wind energy sector, with an additional 10.6 ± 12.5 GW needed—equivalent to a 1.15% ± 1.36% increase compared with normal conditions (Supplementary Fig. 5E).

Conversely, under EPHF scenarios, total energy capacity requirements decline relative to the baseline, though Vietnam remains the largest contributor to regional reductions (Fig. 3C). The bulk of this decrease stems from reduced wind capacity needs, which fall by 9.3 ± 11.7 GW, or 1.02% ± 1.28% relative to normal conditions (Supplementary Fig. 5F). These results highlight Vietnam’s substantial role in shaping LMRB renewable energy pathways, even under extreme climate conditions.

Across the LMRB, changes in total system costs closely mirror shifts in total energy capacity requirements, driven primarily by variations in up-front investment costs. These investment-related changes constitute the majority of system-wide cost fluctuations, outweighing those from fuel and operational expenditures, including both fixed and variable operation and maintenance (O&M) costs (Fig. 3D). Under 100-year EPLF scenarios, total system costs increase by US$5.2 ± 2.6 billion (0.91% ± 0.46%) relative to the baseline mean of US$570.0 billion (Supplementary Table 1). This increase becomes more pronounced under the most severe EPLF conditions; e.g., for 10,000-year EPLFs, additional costs rise to US$8.7 ± 3.2 billion (1.53% ± 0.56%, Fig. 3D and Supplementary Table 1).

Vietnam is responsible for the majority of this cost increase, reflecting its substantial share in capacity expansion under EPLF conditions (Fig. 3C). In the 10,000-year EPLF scenarios, system costs in Vietnam alone rise by US$6.0 ± 2.8 billion—accounting for about 70% of the total cost increase across the LMRB (Fig. 3D; Supplementary Table 1). These findings reinforce Vietnam’s central role in shaping both the capacity and cost dynamics of the regional energy system under prolonged low-flow conditions.

Opposing roles of EPLFs and EPHFs on the pace of VRE deployment

Our results show that EPLFs accelerate, while EPHFs delay, the deployment of VRE capacity under a range of climate extremes (Fig. 4A). Under 100-year EPLF conditions, the increase in VRE capacity unfolds in three distinct phases: (1) a substantial early surge of 4.2 ± 7.4 GW (2.27% ± 4.03% above baseline) by 2030; (2) a more moderate increase of 1.2 ± 8.1 GW (0.18% ± 1.24%) by 2040; and (3) a slight reversal with a decrease of 0.3 ± 8.3 GW (0.07% ± 1.68%) by 2050. In contrast, EPHFs of equivalent severity exhibit the opposite trend: a large initial reduction of 3.5 ± 7.5 GW (1.91% ± 4.13%) by 2030, followed by a smaller decrease of 3.0 ± 8.5 GW (0.45% ± 1.30%) by 2040, and a potential rebound with an increase of 0.4 ± 8.9 GW (0.08% ± 1.69%) by 2050. Despite this late-stage recovery, the cumulative net reduction in VRE capacity relative to the baseline remains.

Fig. 4: Trends in the deployment of variable renewable energy (VRE) capacity influenced by climate extremes.

A Variations in VRE capacity needs for each modelled year (2020, 2030, 2040 and 2050) across all extreme climate shock scenarios relative to baseline scenarios in the LMRB. It should be noted that no new capacity is deployed in 2020 across all climate shock scenarios. 1σ represents one standard deviation of 100 realizations for each climate shock scenario. B Differences in the cumulative changes in VRE capacity needs between 2030, 2040 and the changes in 2050 across all extreme climate shock scenarios. Higher positive values denote more capacity deployed before 2040, indicative of a more evident trend towards earlier deployment of VRE capacity in the LMRB.

We use the difference between early-phase (2030 and 2040) and late-phase (2050) capacity changes as a proxy for deployment pace (Fig. 4B). This analysis reveals that EPLFs consistently lead to earlier VRE deployment, while EPHFs result in delayed rollouts compared to baseline conditions—highlighting the influence of hydroclimatic extremes on the temporal dynamics of renewable energy transitions.

The timing of VRE deployment is strongly influenced by the severity of future changes in wetness. As climate extremes intensify, the magnitude of earlier or delayed deployment becomes more pronounced. This pattern is evident in both EPLF and EPHF scenarios: less severe events (100- and 1000-year return periods) result in smaller changes in capacity needs over time compared with the most extreme scenarios (10,000-year events) (Fig. 4A). This relationship is further supported by early-versus-late deployment differences across climate extremes (Fig. 4B). Under EPLF conditions, the LMRB energy system pivots toward alternative VRE sources to offset hydropower deficits, with this transition becoming more urgent as dry conditions worsen. In contrast, under wetter conditions, surplus hydropower enables greater flexibility, allowing delays in VRE investment while maintaining system reliability.

Discussion

Expanding hydropower within renewable energy systems in the LMRB is increasingly challenging due to the already high level of river exploitation—particularly for hydropower^21,29. Future climatic disruptions are also expected to amplify uncertainty in hydropower generation, raising the risk of cascading impacts across national power systems and interconnected regional electricity grids³⁰. To account for these complexities, we employ the PREP-SHOT model, which is specifically designed to capture the non-linear and dynamic interactions between water and energy systems, thereby incorporating some of these cascading effects into energy planning⁴.

Our simulations demonstrate how extreme climate shifts can substantially influence the timing of VRE capacity investments required to stay on track with decarbonization targets. Drought risk, in particular, underscores the need for adaptive energy planning strategies—such as scaling up system-wide capacity investments—to offset hydropower shortfalls and associated costs³¹. In contrast, EPHFs may offer cost-saving opportunities in energy planning by increasing water availability. When managed effectively, high flows can boost hydropower output³². Greater hydropower output can reduce reliance on VRE sources, lower total system-wide costs, and present a more cost-competitive planning pathway compared to baseline scenarios. Recognizing and planning for this early-versus-late deployment dynamic is critical to ensure long-term energy system resilience under climate uncertainty.

Our comparative analysis of 700 climate shock scenarios highlights the contrasting effects of EPLFs and EPHFs on the timing of VRE deployment in the LMRB. The scenarios demonstrate that strategic, system-level energy planning—requiring relatively modest additional investments in selected countries—can enhance energy system resilience and effectively “climate-proof” infrastructure against future extremes³³. For instance, an additional US$3.5 billion investment over 2020–2050—equivalent to approximately 0.6% of the LMRB’s total GDP in 2023—is sufficient to mitigate the impacts of severe EPLF-induced hydropower reductions, spanning 100- to 10,000-year return periods (Supplementary Table 1). Our simulations further show that EPHF conditions delay VRE deployment by generating temporary surpluses in hydropower, which reduce the urgency of investment in alternative renewables—introducing long-term uncertainties for energy system resilience. While EPHFs may alleviate short-term VRE deployment pressure, our findings also imply potential long-term resilience trade-offs. Increased hydropower availability under EPHFs delays VRE investments, reduces generation diversity, and increases vulnerability to hydro variability. This increased reliance on hydropower may create structural vulnerabilities if future flow variability intensifies or shifts unpredictably. To enhance long-term resilience, planning frameworks should therefore evaluate not only cost-optimal trajectories but also the timing risk of deferring diversification—recognizing that short-term hydropower surpluses may not guarantee future system stability under compounding climate extremes.

To balance the opposing impacts of EPHFs, proactive investment and planning strategies are essential to adapt the power mix to future fluctuations in hydropower generation³². Incorporating these findings, stakeholders can implement targeted strategies to mitigate climate risks, reduce environmental impacts, diversify renewable energy portfolios and enhance cross-border electricity cooperation³⁴. This approach not only strengthens energy system resilience but also ensures alignment with long-term decarbonization goals through adaptive and cost-effective planning in the LMRB. However, realizing the full potential of cross-border cooperation will require overcoming economic, institutional, and political barriers.

We acknowledge that our model may underestimate energy capacity requirements and associated system costs under various climate extreme scenarios. Our approach assumes well-coordinated transboundary electricity transmission through cross-country cooperation, again assuming the barriers for effective cooperation are removed. Disparities in burden-sharing (Fig. 5) and varying susceptibility to climate extremes among countries could challenge such transboundary cooperation (Fig. 5 and Supplementary Table 1). Second, while it is common practice in capacity expansion models, the omission of detailed land-use constraints and additional grid integration costs for VRE development may lead to underestimation of total energy expansion costs.

Fig. 5: Comparison of the system costs by scenario over the entire planning period (2020 to 2050) within the Lower Mekong River Basin (LMRB).

The results for seven different climate shock scenarios, each including 100 ensemble members. Values on the left vertical axis represent the systems costs have been normalized to a range of 0 to 1, where 0 represents the minimum value and 1 represents the maximum value across all scenarios for a specific country. Numbers in brackets on the right vertical axis represent the mean total system costs (billion USD) for each climate shock scenario. Numbers under the horizontal axis represent the average contribution of each country’s cost accounting for the total system cost across all scenarios.

Another limitation concerns the 4-hour temporal resolution of our PREP-SHOT model, which may smooth out, and therefore, underestimate peak electricity load demands²¹. Because peak load demand is a critical constraint in our optimization problem, underestimations can propagate through the energy planning process, resulting in diminished projections of both total energy capacity requirements and investment costs in the optimized expansion pathways³⁵. At the spatial scale, our simulations include only the hydropower plants in the basin that are listed in published dam databases³⁶. Smaller plants do exist but are excluded due to the lack of reliable data on their characteristics³⁷.

Both climate projections and future hydropower development involve substantial uncertainty, particularly at the regional scale. In this study, we deliberately exclude such projections and instead focus on hydroclimatic variability derived from historical observations, using a synthetic streamflow ensemble to explore plausible extreme conditions under a stationary climate and infrastructure assumption. Our quantified impacts of extreme events on energy system planning may also be conservative, as we focus on existing hydropower projects. We do not account for the decommissioning of older hydropower reservoirs once they reach their lifetime expectancies, for example because of sedimentation³⁸. We also do not consider future planned hydropower projects in LMRB because there are uncertainties regarding their construction status and locations owing to political instability, economic stress, and social and environmental impacts. For example, the negative consequences caused by the dam boom in LMRB have been the topic of great international scorn^36,39,40. In response, some governments have re-evaluated their dam construction standards, even suspending some new hydropower projects at a variety of scales¹⁸. We also do not explore the role of hydropower reservoirs on greenhouse gas emissions⁴¹. Therefore, in future studies, it would be valuable to incorporate different hydropower development pathways as dynamic variables, considering the impacts of potential hydropower expansion or decommissioning. Furthermore, we agree that future studies could build on this work by incorporating climate-driven hydrological changes and assessing their implications for energy planning in the region.

Furthermore, innovation in energy storage technologies—particularly grid-scale batteries and emerging hydrogen systems—could profoundly alter energy system dynamics under EPLF scenarios^42,43. As our results show, prolonged droughts reduce hydropower output and drive early investment in wind energy to maintain decarbonization trajectories. However, the temporal mismatch between VRE generation and load demand can increase system-wide capacity needs and associated costs. To aid increased VRE integration under EPLFs, improved energy storage capacity could mitigate these mismatches by absorbing surplus generation and supplying power during shortfalls, thereby reducing the need for overbuilding VRE capacity. For example, cost reductions in lithium-ion batteries^44,45 or advancements in long-duration storage^46,47,48 could enhance grid flexibility and buffer against hydropower volatility induced by climate shocks. In our future work, designed scenarios that incorporate storage innovation and deployment dynamics into energy expansion planning could provide more deep insights into the resilience and affordability of drought-resilient energy transitions.

A few studies have already assessed potential variations in hydropower production under future climate change scenarios in the Mekong region⁴⁹. Uncertainty in future precipitation patterns, which are imperfectly simulated in time and space, leaves open questions regarding whether hydropower generation in this region will increase or decrease^18,50,51. Factoring in these uncertainties can make energy system planning even more difficult⁵². Also pertinent is the extent that river flows will be regulated by a few, large-storage, multi-purpose dams in the upper part of the Mekong basin due to multi-use water diversion and flood control strategies^22,51. These human water management practices are not yet captured in the current version of PREP-SHOT.

Despite inherent technical limitations, our analysis underscores the critical need to integrate climate extremes into basin-scale energy planning to ensure future regional energy security—particularly given the uneven spatial distribution of electricity demand and VRE potential across the LMRB⁵³. For instance, Vietnam is projected to account for 64.9% of the basin’s electricity demand by 2050 (Supplementary Fig. 4), yet its system costs under normal conditions represent only 42.6% of total regional system costs (Supplementary Fig. 6). In contrast, Thailand, with a projected demand share of 25.1%, would bear 33.7% of the system costs (Supplementary Figs. 4, 6).

Moreover, adapting to increasingly severe EPLFs—from 100- to 10,000-year return periods—would require an additional US$3.5 billion in system-wide investment during the planning horizon (Supplementary Table 1). Of this, 60.0% (US$2.1 billion), or approximately 0.5% of Vietnam’s 2023 GDP, would fall to Vietnam (Supplementary Table 1). This uneven distribution of investment burden highlights the challenges of regional coordination and cost-sharing in the face of spatial disparities in energy demand, generation potential, and climate vulnerability (Supplementary Fig. 4; Supplementary Table 1). These findings emphasize the need for coordinated planning frameworks and incentive mechanisms to enable cost-competitive and equitable capacity expansion across the basin⁵⁴. This imbalance also underscores the need for climate finance mechanisms that can support vulnerable countries in managing system-level investment shocks. International climate funds, or regional financing instruments—potentially coordinated through ASEAN or MRC platforms—could be mobilized to subsidize renewable energy investments or grid upgrades in disproportionately affected countries. Equity-based burden-sharing arrangements may also help realign incentives for cross-border cooperation, especially where downstream countries face upstream risks beyond their control. Embedding such financial mechanisms into regional energy planning can enhance both technical robustness and political feasibility of climate-resilient infrastructure transitions.

While this study develops and applies a robust framework for understanding the dynamic interplay between climate extremes and energy system transitions, several critical challenges remain. Economic, institutional, and political barriers continue to hinder cross-border electricity cooperation, limiting the full potential of regional energy integration. Future research should prioritize the development of mechanisms to optimize transboundary electricity trade, improve grid flexibility, and integrate demand-side management strategies to better address energy shortfalls and surpluses under intensifying climate variability. Meanwhile, strengthening transboundary cooperation frameworks such as the MRC could improve the alignment between water governance and energy integration in the LMRB^22,55. As our results demonstrate, extreme flow conditions in hydropower-rich countries such as Laos—where there is substantial untapped potential—can profoundly impact energy security in countries like Vietnam, which despite having high installed capacity, has limited remaining hydropower development potential. During EPLFs, reductions in hydropower exports from Laos exacerbate this vulnerability. Given the increasing reliance on cross-border electricity trade and the interconnected nature of water and energy systems, coordinated reservoir and power system operations are essential. While our analysis assumes fully coordinated electricity transmission across countries, achieving such integration in practice remains challenging due to disparities in climate exposure, and investment burdens. Existing mechanisms like the MRC, though primarily focused on water resources, could serve as a platform to better support integrated water-energy planning. For instance, basin-wide coordination on reservoir releases, hydropower scheduling, and electricity exports could help reduce system-wide costs and improve resilience under climate extremes. However, this would require enhanced data sharing, clearer mandates, and greater political willingness to address the uneven distribution of benefits and risks across countries.

To compensate for the increasing prevalence of VRE, the hydropower operational shift will be adjusted from baseload to a flexibility supplier. Such a shift may damage the river ecology^56,57,58. We did not explicitly incorporate detailed ecological constraints of hydropower operation in the PREP-SHOT model, which may underestimate the cost or overestimate the benefits of hydropower.

Future work could benefit from integrating a multi-objective trade-off framework that explicitly captures the conflicts between system costs, carbon emissions, and ecological outcomes. While our current analysis mainly focuses on optimizing for system-wide cost under climate-induced hydropower variability, decision-makers often face competing priorities—such as minimizing emissions, protecting biodiversity, or distributing investment burdens equitably across countries. For instance, expansion pathways that reduce system-wide cost may come at the expense of river ecosystem health or exacerbate transboundary disparities in climate exposure and responsibility. Embedding such objectives into a unified modeling framework would enable the identification of Pareto-optimal solutions and help stakeholders better understand the trade-offs associated with different development trajectories. This approach could also inform the design of compensation schemes or regulatory safeguards that balance climate resilience with broader sustainability and justice considerations.

More broadly, the modeling framework and insights developed here extend beyond the LMRB, offering valuable guidance for hydropower-dependent regions such as West Africa⁵⁹, the Nile Basin⁶⁰, and the Amazon Basin⁷. For example, in the Nile Basin, studies show that upstream climatic shocks and water use patterns can propagate downstream and reshape hydropower availability and energy security^21,61. As climate change drives more frequent and severe hydrological extremes, energy systems must adopt adaptive, basin-scale strategies that account for coupled water-energy dynamics. Incorporating these extremes into energy planning is critical for ensuring cost-effective, resilient, and sustainable transitions toward low-carbon energy futures^62,63.

Methods

Our analysis mainly includes three parts: (1) simulation of hydroclimate shock scenarios based on synthetic streamflow data; (2) energy expansion modelling forced by the synthetic streamflow data; and (3) evaluation of climate extreme impacts on energy system development.

Synthetic streamflow generation

Hydropower development planning in Southeast Asia is limited by a lack of historical hydrometeorological data and an inherent uncertainty of the frequency and intensity of hydroclimate river flow extremes. In the absence of long-term river flow time series, effective planning necessitates the development of synthetic streamflow series that encompass the realm of extreme flows that would ultimately affect hydroelectricity production, a major component of the current renewable energy portfolio in Southeast Asia^18,36. Statistical synthetic streamflow generators are cost-effective in simulating a comprehensive range of extreme climate shock scenarios, thereby facilitating robust evaluations of the potential impacts of climate extremes on energy systems⁶⁴. It should be noted that “climate shock scenarios” refer to synthetic hydroclimatic conditions generated from statistical models based on historical observations in this study. These are distinct from GCM-derived future climate projections and are intended to explore plausible but extreme realizations of baseline climate variability.

Our study employs a non-parametric statistical approach developed by Kirsch et al.⁶⁵, which not only reproduces key streamflow statistics, such as seasonal means and variances, but also preserves spatial cross-correlations across gauges. Unlike other process-based hydrological models, this generator is driven solely by historical streamflow records without requiring additional data. A number of prior studies have applied this approach to a variety of applications, such as water resources management⁶⁶ and reservoir operations⁶⁷.

In our experiments, we generate synthetic streamflow for four representative years (2020, 2030, 2040, and 2050) for 57 existing large hydropower reservoirs within the LMRB (Fig. 2A). This is done under various hydroclimatic scenarios, each including 100 ensemble members. The detailed workflow includes the following 12 steps:

Step 1: Calculate the monthly streamflow data at the outlet of the LMRB by summing the historical daily values.

Step 2: Standardize monthly streamflow values with a log transformation, then reorganize the data into a matrix $X$ of n rows (years) by 12 columns (months).

Step 3: Compute the cross-month correlation matrix ($C={X}^{T}\cdot X$) using standardized historical streamflow data obtained from Step 2. Perform a Cholesky decomposition ($C=L\cdot {L}^{T}$), for which L is a real lower triangular matrix with positive diagonal entries and ${L}^{T}$ is its transpose.

Step 4: For each month, bootstrap the standardized streamflow data, recording the associated years as a timestamp to ensure that generated synthetic streamflow time series for other locations maintain the spatial-temporal correlation structure across the entire LMRB domain. Repeat this process 10,000 times to construct a matrix (${Z}_{0}$) of random and uncorrelated synthetic streamflow data (10,000 by 12).

Step 5: Impose the temporal autocorrelation structure by multiplying the bootstrapped streamflow ${Z}_{0}$ with the coefficient matrix $L$ from Step 3: ${Z}_{1}={Z}_{0}\cdot L$, where ${Z}_{1}$ is the resampled standardized streamflow constrained by temporal correlations.

Step 6: Concatenate all ${Z}_{1}$ matrices for all four representative years (2020, 2030, 2040, and 2050).

Step 7: De-standardize the data by applying the inverse log transformation to convert the log-transformed, standardized streamflow values back to their original form. Generate 10,000 realizations of four-year synthetic streamflow data in a matrix format of 10,000 by 48.

Step 8: Rank all 10,000 realizations in ascending order based on their four-year total streamflow. Specifically, EPLF and EPHF scenarios are defined based on the ranking of 10,000 ensemble realizations of four-year cumulative streamflow. These realizations are sorted in ascending order, and specific ranked members are selected to represent different return periods of hydroclimatic extremes. For EPLFs, we select the 1st, 10th, and 100th ranked realizations, corresponding approximately to 1-in-10,000, 1-in-1000, and 1-in-100-year extreme dry conditions, respectively. For EPHFs, we use the 9900th, 9990th, and 10,000th ranked realizations to represent symmetric return periods for extreme wet conditions. Additionally, the 5000th realization, which lies at the median of the distribution, is used to represent a normal hydroclimate scenario.

Steps 1-8 generate synthetic streamflow at the LMRB outlet; however, as we are interested in plant-level hydropower production, we generate synthetic streamflow for each individual upstream reservoir in Steps 9-12.

Step 9: Generate synthetic streamflow into reservoirs, maintaining the spatial-temporal correlation structure of inflows to the 57 reservoirs using the recorded timestamp information of each basin outlet in Step 4 to sample synthetic streamflow. Generate synthetic streamflow data for all reservoirs by selecting their respective historical streamflow data from the same years and months corresponding to the selected extreme and normal scenarios from Step 8.

Step 10: Disaggregate the synthetic monthly streamflow data for all reservoirs into daily data based on their historical observations at daily time scales using the K-Nearest Neighbour (KNN) method⁶⁸.

Step 11: Repeat Steps 4-10 100 times to construct the ensembles of synthetic streamflow time series.

Step 12: Further disaggregate all daily streamflow data into a four-hourly interval by sampling the daily values repeatedly. This final step prepares the streamflow data for compatibility with the energy capacity expansion model (see details in the next section).

The above steps allow the generation of synthetic streamflow data for four representative years (2020, 2030, 2040, and 2050) and seven climate shock scenarios. These scenarios include one baseline (normal conditions), three wet scenarios (EPHF conditions with return periods of 100, 1000, and 10,000 years), and three dry scenarios (EPLF conditions with return periods of 100, 1000, and 10,000 years). Each scenario comprises 100 ensemble members for 57 selected reservoirs.

Preliminary analysis shows that the synthetic generation algorithm can competently generate streamflow of each individual power plant (Supplementary Table 2). All ensembles of streamflow data generated by the synthetic generation algorithms maintain the same dependencies (spatiotemporal dynamics and cross-correlations among reservoirs on seasonal time scales) and statistical moments as the historical record (see details in Supplementary Fig. 7 and Supplementary Table 2). Coupled with an energy capacity expansion model, this ensemble of synthetic streamflow data enables a quantitative assessment of the impacts of climate extremes on the optimal energy expansion pathways in the LMRB.

Energy capacity expansion model

To identify the optimal energy expansion pathway of LMRB under various climate shock scenarios, we use a modular and open-source energy expansion model: PREP-SHOT⁴. Compared with other hydropower-centric energy planning models^2,5, the unique feature of PREP-SHOT is that it hard-couples a multi-reservoir system model and a power system model. This integration captures two-way feedback between short-term hydropower operation and long-term energy system planning and thereby makes more realistic operational decisions upon the states of both the water and energy systems. This capability is especially critical in regions with a large number of cascade hydropower stations, for example in the LMRB. Details of PREP-SHOT are documented in Liu & He (2023)⁴. The code of the stable version can be accessed at: https://prep-next.github.io/PREP-SHOT/.

Major inputs of PREP-SHOT include reservoir inflows, existing power infrastructures (such as power plant types, transmission lines, and energy storage; Supplementary Tables 3 and 4), and the capacity factors of VRE. It also incorporates projected electricity load demand over the planning horizon, decarbonization targets, and various techno-economic parameters. These parameters include the lifetime of power technologies and transmission lines, ramping rates for power technologies, lower and upper bounds of the country-level installed capacities of each technology, carbon emission factors of thermal power plants, and electricity transmission topology and efficiency (Supplementary Tables 5 and 6). Additionally, economic factors for each country in the LMRB are considered, including discount rate, investment cost, fixed O&M cost, variable O&M cost, and fuel cost.

All energy technologies included in PREP-SHOT are classified into four categories: (1) ‘hydro’; (2) ‘storage’; (3) ‘non-dispatchable’; and (4) ‘dispatchable’. PREP-SHOT incorporates a ‘hydro’ generation process within specific locations at the plant level. Notably, hydropower in PREP-SHOT not only serves as an important component of electricity generation to meet load demand, but also provides flexibility services to support renewable integration, especially in future energy systems with a high penetration level of VRE. For ‘hydro’, we use synthetic reservoir inflows at four-hour time scales for each modelled year to estimate plant-level hydropower generation. For ‘storage’ technologies, two currently available energy storage technologies, pumped storage hydropower (PSH) and lithium-ion (Li-ion) battery, are considered. PREP-SHOT determines which energy storage type is most cost-effective in facilitating renewable integration, when these energy storage technologies are deployed, and the corresponding capacity of energy storage technologies installed in each country. ‘Non-dispatchable’ technologies consist of solar and wind energy, both of which are limited by capacity factors driven by local weather conditions and installed capacity. ‘Dispatchable’ technologies, including coal-fired plants, oil-fired plants, gas-fired plants and bioenergy, can be controlled within a certain range and usually serve as complementary and flexible power supplies.

Resilient and robust planning of power grids must consider the variably fluctuating energy supply, for example, daily variation of solar and wind energy, intra-annual variability of hydropower generation. This aspect is highly relevant when extreme events are explicitly considered. The energy capacity expansion model must therefore effectively capture the fluctuations in energy supply to ensure that the amount of power generated (electricity supply) is closely balanced with the amount of power being consumed (electricity demand). Traditional models routinely do this by selecting representative days, rather than considering the entire year, to avoid resource-prohibitive computation⁶⁹. However, this type of short-cut may fail to accurately capture the intra-annual variability of load demand especially when the extreme days of the net load are not included, resulting in an underestimation of the required dispatchable capacity and investment costs when making planning and operational decisions. To avoid this issue, we run four full-year (i.e., representative years 2020, 2030, 2040, and 2050) simulations using PREP-SHOT at four-hourly temporal timesteps (a total of 2,190 = 365*6 time intervals for each modelled year). This solution not only can accurately capture the intra-annual variability of load demand when optimizing the optimal energy expansion pathways but can balance the trade-offs between model representability and computational efficiency. The objective function minimizes the total cost of electricity power planning and operation.

700 synthetic streamflow scenarios (i.e., 100*7 ensembles of four-hourly streamflow time series for each modelled year) are used as inputs for PREP-SHOT to characterize the severities and uncertainties of wetness extremes under different climate shock scenarios. Regarding the spatial resolution, we select five LMRB’s countries as “spatial nodes”: Cambodia, Laos, Thailand, Vietnam, and Myanmar. As the locations of detailed electrical substations in these countries are not publicly available, we model the power transmission lines across different countries as direct connections between the capital of each country (Fig. 2B). Existing and newly installed capacities of all technologies are allocated to each country, as well as future load demand and renewable energy generation time series. Electricity transfer within these five LMRB countries is allowed on the assumption that full coordination can be achieved through the operation of the power grid. The total amount of electricity generation from different energy types and electricity imports (or exports) is restricted to meet the electricity load demand at four-hourly time scales.

PREP-SHOT outputs three major types of variables: water, energy, and cost. Variables related to water systems include the flow released over spillways, as well as storage dynamics for all reservoirs during each time interval. Energy system variables include newly installed capacities during the planning period including different technologies and transmission lines for each modelled year at the country level. These variables are obtained from the cost-optimal pathways using PREP-SHOT. Meanwhile, hourly electricity generation of each technology and transmitted power between paired neighbouring countries per modelled year is also provided to characterize the electricity export or import among each country in LMRB. Cost-related variables include fuel expenses, variable and fixed O&M costs, as well as annualized investment costs for each technology and transmission lines for each modelled year within each country.

Data assumptions and explanations

Natural inflow data

Due to the limited availability of observation-based reservoir inflow data, we use simulated natural inflow obtained from Xu & He (2022)³⁸ to generate synthetic streamflow of all selected reservoirs used in this study (Fig. 2A). Xu and He³⁸ leveraged the calibrated Soil and Water Assessment Tool (SWAT) model to simulate the daily natural inflow of these reservoirs from 1962 to 2005. Furthermore, daily streamflow observations at Kratie station (105.45° E, 12° N, marked in Fig. 2A) over 1962-2005, provided by the MRC, are collected to reflect the regional hydroclimate conditions in LMRB, as it is near to the basin outlet of LMRB (see the section on “Synthetic streamflow generation” for details).

Hydropower generation

Detailed information of plant-level reservoir characteristics and cascade topologies of selected hydropower stations allows capturing realistic hydraulic connections for watersheds with cascade reservoirs. Doing so is essential to accurately simulate the plant-level hydropower generation process especially in regions with a large number of cascade hydropower stations, including some catchments in the LMRB. In this study, we apply a constant water travel time-based routing method to simulate hydraulic delays across cascade systems. This method, commonly used in short-term hydropower operations (Liao et al.)⁷⁰, effectively captures the timing of water transfers between reservoirs while maintaining computational efficiency in large-scale simulations⁴. The water travel times used in our model are estimated based on a well-calibrated SWAT model from previous research³⁸, ensuring consistency with the hydrodynamic characteristics of the basin. After translating synthetic streamflow series associated with other relevant system variables to model inputs, PREP-SHOT provides the outputs of hydropower generation at each reservoir at a four-hour resolution, which can characterize the influence of climate extremes on water and energy systems better than aggregating all hydropower generation at country-level with a lower temporal scale. It should be noted that due to institutional restrictions and limited data availability, especially for cross-border or privately operated facilities, we were unable to obtain sufficiently detailed records for validation. Therefore, it is challenging to directly compare simulated hydropower generation against observed data.

Solar and wind energy

We use gridded meteorological variables obtained from Modern-Era Retrospective analysis for Research and Application Version-2 (MERRA-2) reanalysis product⁷¹ to estimate the capacity factors of solar and wind power. We select 1986 as the representative year to depict the hourly variations of solar and wind power in each modelled year during 2030-2050, because it can reflect the actual variations of capacity factors in a median year. Furthermore, all hourly values are converted into a four-hourly interval by averaging these data over each four-hourly period with the goal of keeping consistent with the temporal resolution of load demand.

We follow the approach in Liu and He⁴ to estimate the capacity factors of VRE. For solar power, gridded hourly surface incoming shortwave radiation, top of the atmosphere incoming shortwave radiation, and 2-meter temperature are used to calculate pixel-level capacity factors. The capacity factors of wind energy are calculated using hourly 10- and 50-m wind speeds at each pixel. We aggregate pixel-level capacity factors to country-level by spatially averaging all grid cells within a country, weighted by the pixel area.

In addition to the capacity factors of VRE, we also consider the upper limit of solar energy, which constrains how much solar photovoltaic energy can be produced. This is done by setting different upper bounds in PREP-SHOT for different energy types. The upper bound of solar available in each country of LMRB is based mainly on the approach of Siala et al.⁷², because the calculation considers topographic and land-use constraints. An Asian Development Bank (ADB) report⁵⁴ suggests a large potential of wind energy (both onshore and offshore) in LMRB given the vast land and marine areas that are suitable for wind installations. We therefore do not set the technical constraints for wind in PREP-SHOT.

Existing technology capacities and load demand profiles

We collect the 2020 capacity of existing technologies in LMRB, including coal, oil, gas, hydropower, solar, wind, bioenergy, and transmission lines (Fig. 2B) based on published statistics by the Association of Southeast Asian Nations Centre (ASEAN) for Energy⁷³. The capacity and distribution of existing cross-border power transmission lines are acquired from Li & Chang (2015)⁷⁴. To better depict the load demand profiles in LMRB, we collect country-level hourly electricity data in 2020 from published studies^6,19 (see details in Supplementary Note 1). Afterwards, original values are aggregated from hourly to four-hourly intervals to represent the load demand profiles in each modelled year. This aggregation is necessary because it is challenging and computationally expensive to optimize the high-dimensional energy system over an entire year (8,760 h) spanning the full planning horizon that often extends for several decades. In addition, electricity load demand is projected to grow rapidly in the coming decades due to the economic growth in LMRB²⁴. Therefore, we further calculate the four-hourly load demand for each modelled year in the future (2020~2050) by using the projected annual average growth rate of 2.7% for Thailand⁷⁵, 6.0% for Vietnam⁷⁶, 8.8% for Cambodia¹⁹, 9.5% for Laos¹⁹ and 2.7% for Myanmar⁷⁵ with electricity load demand data in 2020 for each country.

Carbon emission limits

For each modelled year between 2020 and 2050, we collect country-level decarbonization targets from the Climate Action Tracker⁷⁷, which rates countries based on their efforts to keep temperature increases well below 2 °C and strive for a limit of 1.5 °C above pre-industrial levels. In this study, we use the CAT ratings for the 1.5 °C limit to establish the maximum allowable carbon emissions for each country, ensuring their policies align with the Paris Agreement’s 1.5 °C goal. We then aggregate these carbon emissions across all countries in the LMRB and apply a unified carbon emission constraint that supports the target of limiting long-term warming to 1.5 °C for the entire region.