Observing glacier mass changes is essential for understanding and projecting the impacts of climate change on sea-level rise, water resources and natural hazards, as well as providing data for developing, calibrating and validating glacier evolution models. The principal methods used to measure glacier mass changes - glaciological, geodetic (surface elevation differencing) and gravimetric - differ in the spatial and temporal scales at which they are most effectively applied. Here, we review these methods in the context of challenges that arise when comparing published mass-balance estimates. Compatibility can be hampered by (1) inconsistent reporting and lack of relevant information; (2) discrepancies in which mass-balance components are included; (3) differences in the time span analyzed; and (4) variations in the spatial domain of the reported mass balance. We provide recommendations for more rigorous and comprehensive reporting of mass-balance estimates to improve comparability and synthesis of reported glacier mass changes, and we emphasize open data and code sharing to enable full reproducibility and future reinterpretation. Our recommendations apply equally to both glacier and ice-sheet mass-balance reporting, and they are generally valid for mass balances simulated by numerical models.

Glaciers outside of the ice sheets are known to be important contributors to sea level rise. In this work, we provide an overview of changes in the mass of the world's glaciers, excluding those in Greenland and Antarctica, between 2002 and 2016, based on satellite gravimetry observations of the Gravity Recovery and Climate Experiment (GRACE). Glaciers lost mass at a rate of 199 ± 32 Gt yr ⁻¹ during this 14-yr period, equivalent to a cumulative sea level contribution of 8 mm. We present annual mass balances for 17 glacier regions, that show a qualitatively good agreement with published estimates from in situ observations. We find that annual mass balance varies considerably from year to year, which can in part be attributed to changes in the large-scale circulation of the atmosphere. These variations, combined with the relatively short observational record, hamper the detection of acceleration of glacier mass loss. Our study highlights the need for continued observations of the Earth's glacierized regions.