Addressing climate uncertainty and incomplete information in transboundary river treaties: A scenario-neutral dimensionality reduction approach

Abstract

• We enumerate 25,000 + solutions for the Ganges water sharing treaty, renewed in 2026. • Hierarchical clustering identifies redundant objectives and governing trade offs. • Pareto-optimization & scenario-neutral analysis finds climate-robust treaty solution. Climate change will alter the flow availability and expected water allocations in international river treaties, many of which were designed using historical flow records. Effective transboundary treaties should anticipate these concerns and seek to satisfy the priorities of all riparian countries while being robust to impending changes in climate. This task is complicated by the fact that specific objectives associated with each party’s priorities are not necessarily common knowledge (framing uncertainty), and the direction, amplitude and effect of long term changes in hydro-climatic drivers can be highly uncertain (climate uncertainty). We frame the design of a transboundary treaty as a multi-objective optimization problem. We use hierarchical clustering to address problem-framing uncertainty by identifying the subset of objectives associated with the governing trade-offs imposed by the bio-physical characteristics of the shared river system. We then carry out a scenario-neutral climate sensitivity analysis to identify climate-robust Pareto-optimal treaty solutions. We illustrate the approach for the Ganges water agreement, which is due to be renewed in 2026. Based on an enumerated population of 25,121 feasible treaty solutions, we identify governing objectives and 16 treaty solutions that are Pareto optimal under most considered combinations of changes in sea level and dry season flow regime. This work provides a path towards improving transboundary allocations for the Ganges water treaty and, more broadly, a template to support transboundary cooperation over shared international rivers.