Multidecadal estimation of hydrological contribution and glacier mass balance in the semi-arid Andes based on physically based modeling and geodetic mass balance

Alonso Mejías ¹ James McPhee ^2* Hazem Mahmoud ³ David Farías-Barahona ⁴ Christophe Kinnard ⁵ Shelley MacDonell ⁶ Santiago Montserrat ⁷ Marcelo Somos-Valenzuela ⁸ Alfonso Fernandez ⁹

• ¹ Department of Civil Engineering, Faculty of Physical Sciences and Mathematics, University of Chile, Santiago, Santiago Metropolitan Region (RM), Chile
• ² University of Chile, Santiago, Chile
• ³ University of Texas at San Antonio, San Antonio, Texas, United States
• ⁴ University of Concepcion, Concepción, VIII Biobío Region, Chile
• ⁵ Département des sciences de l'environnement, Université du Québec à Trois-Rivières, Trois-Rivières, Quebec, Canada
• ⁶ Center for Advanced Research in Arid Zones, Faculty of Marine Sciences, Catholic University of the North, Coquimbo, Chile
• ⁷ Advanced Mining Technology Center, Faculty of Physical Sciences and Mathematics, University of Chile, Santiago, Santiago Metropolitan Region (RM), Chile
• ⁸ Faculty of Agricultural Sciences and Environment, University of La Frontera, Temuco, Chile
• ⁹ Faculty of Architecture, Urban Planning and Geography, University of Concepcion, Concepcion, VIII Biobío Region, Chile

Glaciers are of paramount importance in diverse environments, and due to the accelerated retreat experienced in recent decades, efforts have intensified to achieve a comprehensive understanding of key variables such as mass balance and glacial melting. However, the scarcity of data in regions that are difficult to access, such as the Andes Cordillera, hinders reliable glaciological studies of the historical period. This research examines the mass balance and melting dynamics of the Universidad Glacier, the largest in the semiarid Andes, from 1955 to 2020, using the physically based Cold Regions Hydrological Model (CRHM). The model was calibrated with geodetic mass balance estimates available between 1955 to 2020 and evaluated against on-site observations available between 2012 to 2014. Change point analysis revealed three contrasting periods of mass balance evolution: significant mass loss for the periods 1955-1971 and 2006-2020, and near equilibrium mass balance from 1971-2006. These loss and gain periods align with negative phases of the Pacific Decadal Oscillation (PDO) and positive ENSO (El Niño) events, respectively. Simulated runoff from glacier melt showed a positive trend of 8% per decade since 1971. Calibrated and uncalibrated versions of the model showed similar temporal variability, but cumulative mass balance differed significantly. The model calibrated from 1955 to 2020 had a minimal overestimation of 0.1% in mass loss and slightly improved the representation of annual albedo. Relative to this best-performing model, the model calibrated with geodetic mass balance estimates from 2000 to 2020 overestimated mass loss by 25%, while the uncalibrated model overestimated mass balance by 62%. Physically based modeling with parameters adjusted based on field observations is adequate to reproduce the most salient features of MB interannual variability. However, long-term projections may diverge significantly, and albedo parameterizations, including its special and temporal evolution throughout a glacier surface, are an avenue for future research.