In arid-zone water resource management and water-security assessment, changes in water-body volume are key indicators of water availability and regulation performance. However, arid-zone lakes often lack sufficient bathymetric information to constrain geometry under low lake-level conditions. Shrinkage-driven hydrological disconnection can destabilize extrapolation of water level–storage relationships. This increases uncertainty in quantifying long-term storage changes. Here, we develop a multi-digital elevation model (DEM) hypsometry framework to reconstruct near-monthly lake storage for 1993–2024, recovering storage during low-level periods without bathymetric surveys. Reconstructed changes agree with independent satellite altimetry (r = 0.93 for level and 0.90 for storage), outperforming above-water-only (r ≈ 0.637 for water level) and conventional model-selection base-lines (r ≈ 0.753 for water level). The framework was quantified across three scenarios: expanding lakes, lake systems and reservoirs, and terminally shrinking lakes. For the persistently shrinking Big Aral Sea, under the whole-lake modeling assumption, the Copernicus-based reconstruction provides a cumulative storage change of −214.3 km3, closest to the satellite altimetry estimate of −210.68 km3 among the tested DEMs. In contrast, other DEMs overestimate the 1993–2024 cumulative loss by 66.15–141.01 km3. Sub-lake modeling further adjusts the Shuttle Radar Topography Mission (SRTM)-based cumulative change to −248.38 km3, substantially reducing structural bias caused by lake disconnection. This study provides a transferable technical framework for lake storage reconstruction in arid regions under degraded low lake-level conditions and hydrological disconnection.
Huang et al. (2026) studied this question.