Glacier mass balance (MB) is central to understanding glacier change and calibrating glacier evolution models, yet direct MB observations remain sparce. Recent efforts to infer distributed MB fields from the mass continuity equation have so far mainly been limited to individual glaciers, used simplified elevation-dependent approaches, and/or relied on glacier models for the inversion. Here, we present one of the first regional applications of a continuity-based inversion to derive spatially distributed MB fields directly from remote sensing observations of ice thickness, surface velocity, and surface elevation change, for all glaciers larger than 2 km² in the Swiss Alps.</p> <p>Using in-situ MB observations from the Swiss Glacier Monitoring Network, we optimize and evaluate our inversion framework for regional performance and transferability. We show that numerical filtering of the ice flux divergence is essential to obtain physically meaningful MB fields. After filtering, the inferred MB fields closely match observations (RMSE = 0.76 m w.e. yr⁻¹). Sensitivity analyses indicate robust performance across different ice thickness products, including for global reconstructions that do not integrate local ice thickness observations, while emphasizing the importance of temporal consistency between the input datasets.</p> <p>Our results demonstrate that observation-based distributed MB fields can be derived at regional scale from widely available remote sensing products, offering the method a strong potential for application in other regions worldwide. The resulting MB fields, together with derived equilibrium line altitudes and MB gradients, provide new and independent constraints for glacier model calibration.