Can Earth Observation and AI Reinforce—or Flatten—the Global Hydrological Divide?

Recent work has warned that Earth observation (EO) and artificial intelligence (AI) may reinforce hydrological inequality because satellite infrastructure, computing, validation capacity, and scientific institutions remain unevenly distributed (Di Baldassarre et al., 2026). We agree, but argue that the same technologies can also narrow the divide. Not intrinsic causes of inequality, EO and AI are best viewed as technological multipliers: their effects depend on openness, transferability, validation, and local capability. The appropriate benchmark is therefore an evolving hydrological counterfactual—relative to feasible alternatives, do these technologies reduce or reinforce the disadvantages imposed by geography and institutional wealth? We call a sustained reduction in those disadvantages flattening the global hydrological divide. Open EO can partially decouple access to water-cycle observations from ownership of dense national monitoring infrastructure, while AI can lower analytical barriers when its gains are transferable, uncertainty-aware, testable, and locally challengeable. We organize this argument around three linked capabilities— observe, infer, and act —and propose corresponding tests for whether benefits reach sparsely observed and resource-constrained regions. Hydrology thus provides a revealing test case for a broader Earth science question: whether global observing and AI infrastructure distributes scientific capability or deepens existing gaps.