Jordi Bolibar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Jordi Bolibar has a total of 88 co-authors. Here are the top co-authors by number of shared publications:
- Isabelle Gouttevin
Université Grenoble Alpes, Université de Toulouse, Météo-France, CNRS, CNRM, Centre d’Études de la Neige, Grenoble, France
10 shared publications Nicolas Champollion
Univ. Grenoble Alpes, CNRS, IRD, G-INP, Institut des Géosciences de l’Environnement, Grenoble, France
9 shared publications- Antoine Rabatel
Université Grenoble Alpes, CNRS, IRD, INRAE, Grenoble-INP, Institut des Géosciences de l’Environnement, Grenoble, France
8 shared publications Clovis Galiez
Univ. Grenoble Alpes, CNRS, G-INP, Laboratoire Jean Kuntzmann, Grenoble, France
8 shared publications- Lucille Gimenes
Institut des géosciences de l’environnement, Université Grenoble-Alpes, Grenoble, France
8 shared publications - Romain Millan
Université Grenoble Alpes, CNRS, INRAE, IRD, Grenoble INP, IGE, Grenoble, France
8 shared publications
Latest Publications
- Open Access
- 1 October 2026
- Published by Copernicus GmbH
- Vol. 2026, 1-37
Comment on egusphere-2026-3523
The glacierized mountains are impacted by dramatic changes in the context of global warming with large implications for hydrology. However, the projection of icemelt contribution to streamflow remains a complicated task and is subject of large uncertainties, especially when the calibration-validation process was performed with little or no cryospheric data. In this study, we present the integration of a glacier module in the distributed conceptual hydrological model J2000, taking into consideration the evolution of glacier surface in the simulations. A calibration and a validation was done using a large panel of hydro-cryospheric datasets, including glacier mass balance observations and simulations, snow reanalyses, snow-cover by satellites and stream gauges. The calibration-validation was performed by applying J2000-Rhône-glaciers to the entire Rhône basin in the historical period (1976–2022) using the SAFRAN climate reanalysis at 8 km horizontal resolution. The set of parameters identified in the past and meteorological data from 5 “GCM-RCM” chains of climate projections under RCP 8.5 scenario were used to simulate the hydrological response of the Rhône basin under climate change projections (1976–2095). The results show that snowmelt, icemelt and streamflow are satisfactorily simulated by J2000-Rhône-glaciers in the Rhône basin. The projections show a decrease of annual streamflow expected by the end of the 21st century due to a decrease of precipitation and glacier shrinkage in highly glaciated subbasins. The results of the simulations also show an increase of streamflow in winter but a dramatic decrease in summer mainly due to a change in snow seasonality. The icemelt will become extremely low in Isère, Drac and Durance subbasins, while in the Arve, upper Rhône, and along the Rhône River, the future evolution of icemelt contribution to summer streamflow is very uncertain. Despite these uncertainties, our results suggest that the remaining glaciers will still be crucial to sustain the summer streamflow at the end of the century. However, the peak water may be passed by 2040s in the Arve subbasin, and 2050s in the upper Rhône, and the change in snow seasonality will still threaten the availability of water resources in most of the Rhône subbasins. The use of multi-source cryospheric datasets allowed to gain confidence in the projections of icemelt contribution to streamflow in a large basin in the Alps.
- 1 October 2026
- Published by Copernicus GmbH
- Vol. 2026, 1-37
Open AccessComment on egusphere-2026-3523
The glacierized mountains are impacted by dramatic changes in the context of global warming with large implications for hydrology. However, the projection of icemelt contribution to streamflow remains a complicated task and is subject of large uncertainties, especially when the calibration-validation process was performed with little or no cryospheric data. In this study, we present the integration of a glacier module in the distributed conceptual hydrological model J2000, taking into consideration the evolution of glacier surface in the simulations. A calibration and a validation was done using a large panel of hydro-cryospheric datasets, including glacier mass balance observations and simulations, snow reanalyses, snow-cover by satellites and stream gauges. The calibration-validation was performed by applying J2000-Rhône-glaciers to the entire Rhône basin in the historical period (1976–2022) using the SAFRAN climate reanalysis at 8 km horizontal resolution. The set of parameters identified in the past and meteorological data from 5 “GCM-RCM” chains of climate projections under RCP 8.5 scenario were used to simulate the hydrological response of the Rhône basin under climate change projections (1976–2095). The results show that snowmelt, icemelt and streamflow are satisfactorily simulated by J2000-Rhône-glaciers in the Rhône basin. The projections show a decrease of annual streamflow expected by the end of the 21st century due to a decrease of precipitation and glacier shrinkage in highly glaciated subbasins. The results of the simulations also show an increase of streamflow in winter but a dramatic decrease in summer mainly due to a change in snow seasonality. The icemelt will become extremely low in Isère, Drac and Durance subbasins, while in the Arve, upper Rhône, and along the Rhône River, the future evolution of icemelt contribution to summer streamflow is very uncertain. Despite these uncertainties, our results suggest that the remaining glaciers will still be crucial to sustain the summer streamflow at the end of the century. However, the peak water may be passed by 2040s in the Arve subbasin, and 2050s in the upper Rhône, and the change in snow seasonality will still threaten the availability of water resources in most of the Rhône subbasins. The use of multi-source cryospheric datasets allowed to gain confidence in the projections of icemelt contribution to streamflow in a large basin in the Alps.
- Open Access
- 17 August 2026
- Published by Copernicus GmbH in EGUsphere
The use of multi-source cryospheric datasets to improve the simulation of future icemelt contribution to streamflow in large glaciarized basins
The glacierized mountains are impacted by dramatic changes in the context of global warming with large implications for hydrology. However, the projection of icemelt contribution to streamflow remains a complicated task and is subject of large uncertainties, especially when the calibration-validation process was performed with little or no cryospheric data. In this study, we present the integration of a glacier module in the distributed conceptual hydrological model J2000, taking into consideration the evolution of glacier surface in the simulations. A calibration and a validation was done using a large panel of hydro-cryospheric datasets, including glacier mass balance observations and simulations, snow reanalyses, snow-cover by satellites and stream gauges. The calibration-validation was performed by applying J2000-Rhône-glaciers to the entire Rhône basin in the historical period (1976–2022) using the SAFRAN climate reanalysis at 8 km horizontal resolution. The set of parameters identified in the past and meteorological data from 5 “GCM-RCM” chains of climate projections under RCP 8.5 scenario were used to simulate the hydrological response of the Rhône basin under climate change projections (1976–2095). The results show that snowmelt, icemelt and streamflow are satisfactorily simulated by J2000-Rhône-glaciers in the Rhône basin. The projections show a decrease of annual streamflow expected by the end of the 21st century due to a decrease of precipitation and glacier shrinkage in highly glaciated subbasins. The results of the simulations also show an increase of streamflow in winter but a dramatic decrease in summer mainly due to a change in snow seasonality. The icemelt will become extremely low in Isère, Drac and Durance subbasins, while in the Arve, upper Rhône, and along the Rhône River, the future evolution of icemelt contribution to summer streamflow is very uncertain. Despite these uncertainties, our results suggest that the remaining glaciers will still be crucial to sustain the summer streamflow at the end of the century. However, the peak water may be passed by 2040s in the Arve subbasin, and 2050s in the upper Rhône, and the change in snow seasonality will still threaten the availability of water resources in most of the Rhône subbasins. The use of multi-source cryospheric datasets allowed to gain confidence in the projections of icemelt contribution to streamflow in a large basin in the Alps.
- 17 August 2026
- Published by Copernicus GmbH in EGUsphere
Open AccessThe use of multi-source cryospheric datasets to improve the simulation of future icemelt contribution to streamflow in large glaciarized basins
The glacierized mountains are impacted by dramatic changes in the context of global warming with large implications for hydrology. However, the projection of icemelt contribution to streamflow remains a complicated task and is subject of large uncertainties, especially when the calibration-validation process was performed with little or no cryospheric data. In this study, we present the integration of a glacier module in the distributed conceptual hydrological model J2000, taking into consideration the evolution of glacier surface in the simulations. A calibration and a validation was done using a large panel of hydro-cryospheric datasets, including glacier mass balance observations and simulations, snow reanalyses, snow-cover by satellites and stream gauges. The calibration-validation was performed by applying J2000-Rhône-glaciers to the entire Rhône basin in the historical period (1976–2022) using the SAFRAN climate reanalysis at 8 km horizontal resolution. The set of parameters identified in the past and meteorological data from 5 “GCM-RCM” chains of climate projections under RCP 8.5 scenario were used to simulate the hydrological response of the Rhône basin under climate change projections (1976–2095). The results show that snowmelt, icemelt and streamflow are satisfactorily simulated by J2000-Rhône-glaciers in the Rhône basin. The projections show a decrease of annual streamflow expected by the end of the 21st century due to a decrease of precipitation and glacier shrinkage in highly glaciated subbasins. The results of the simulations also show an increase of streamflow in winter but a dramatic decrease in summer mainly due to a change in snow seasonality. The icemelt will become extremely low in Isère, Drac and Durance subbasins, while in the Arve, upper Rhône, and along the Rhône River, the future evolution of icemelt contribution to summer streamflow is very uncertain. Despite these uncertainties, our results suggest that the remaining glaciers will still be crucial to sustain the summer streamflow at the end of the century. However, the peak water may be passed by 2040s in the Arve subbasin, and 2050s in the upper Rhône, and the change in snow seasonality will still threaten the availability of water resources in most of the Rhône subbasins. The use of multi-source cryospheric datasets allowed to gain confidence in the projections of icemelt contribution to streamflow in a large basin in the Alps.
- 12 August 2026
- Published by Taylor & Francis in The Polar Journal
- p. 1-7
TROPIPOLAR – GLASCLIM project: advancing scientific collaboration on glacier mass balance modelling across the Andes and the Antarctic Peninsula through machine learning
Glaciers worldwide are undergoing rapid retreat and thinning (Zemp, Huss, Thibert, Eckert, McNabb, Huber, Barandun, et al. 2019), driving major environmental impacts across scales. While in-situ an...
- 12 August 2026
- Published by Taylor & Francis in The Polar Journal
- p. 1-7
TROPIPOLAR – GLASCLIM project: advancing scientific collaboration on glacier mass balance modelling across the Andes and the Antarctic Peninsula through machine learning
Glaciers worldwide are undergoing rapid retreat and thinning (Zemp, Huss, Thibert, Eckert, McNabb, Huber, Barandun, et al. 2019), driving major environmental impacts across scales. While in-situ an...
- 4 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Past and future evolution of French Alpine glaciers in a changing climate : a deep learning glacio-hydrological modelling approach
Evolution passée et future des glaciers alpins français sous changement climatique : une approche de modélisation glacio-hydrologique avec deep learning: Les Alpes européennes sont parmi les régions du monde les plus touchées par le changement climatique, avec des taux de recul des glaciers parmi les plus élevés. Les interactions à long terme entre la société, les écosystèmes de montagne et les glaciers de la région soulèvent d'importantes questions sur l'évolution future des glaciers et les impacts environnementaux et socio-économiques qui en découlent. Afin d'évaluer correctement la réponse régionale des glaciers des Alpes françaises au changement climatique, il est nécessaire de disposer d'outils de modélisation adéquats. Dans ce travail, nous explorons de nouvelles façons d'aborder à la fois l'évolution des glaciers et la modélisation glacio-hydrologique à l'échelle régionale. La modélisation de l'évolution des glaciers a traditionnellement été réalisée à l'aide d'approches empiriques ou physiques, dont l'optimisation est de plus en plus difficile compte tenu de la quantité croissante de données disponibles. Ici, nous présentons, à notre connaissance, le premier effort jamais entrepris pour appliquer l'apprentissage profond (i.e. des réseaux neuronaux artificiels profonds) pour simuler l'évolution des glaciers. Comme les système climatique et glaciaire sont tous deux fortement non linéaires, les modèles traditionnels linéaire de bilan de masse offrent une représentation limitée des interactions entre le climat et les glaciers. Nous montrons comment des non-linéarités importantes liées au bilan de masse des glaciers sont capturées par une méthode d'apprentissage profond, ce qui améliore considérablement les performances des modèles par rapport aux méthodes linéaires.Cette nouvelle méthode a été appliquée pour la première fois dans une étude visant à reconstruire les changements annuels du bilan de masse de tous les glaciers des Alpes françaises pour la période 1967-2015. En utilisant des réanalyses climatiques, des données topographiques et des inventaires de glaciers, nous démontrons comment une telle approche peut être utilisée avec succès pour reconstruire les changements de bilan de masse à grande échelle à partir d'observations. Cette étude a également apporté de nouveaux éclairages sur l'évolution des glaciers dans les Alpes françaises au cours du dernier demi-siècle, confirmant les taux de bilan de masse observés plutôt neutres dans les années 1980 et montrant une accélération bien marquée de la perte de masse à partir des années 2000. On constate des différences importantes entre les régions, le massif du Mont-Blanc présentant la perte de masse la plus faible et le Chablais étant le plus touché. Ensuite, nous avons appliqué ce cadre de modélisation pour simuler l'évolution future de tous les glaciers de la région selon de multiples scénarios de changement climatique (N=29). Nos estimations indiquent que la plupart du volume de glace dans la région sera perdue d'ici la fin du XXIe siècle, indépendamment des scénarios climatiques futurs. Nous prévoyons des pertes moyennes de volume des glaciers de 74%, 80% et 88% dans le cadre des scénarios RCP 2.6 (n=3), RCP 4.5 (n=13) et RCP 8.5 (n=13), respectivement. D'ici la fin du XXIe siècle, les Alpes françaises seront en grande partie libres de glace, avec des glaciers ne subsistant que dans les massifs du Mont-Blanc et du Pelvoux.
- 4 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Past and future evolution of French Alpine glaciers in a changing climate : a deep learning glacio-hydrological modelling approach
Evolution passée et future des glaciers alpins français sous changement climatique : une approche de modélisation glacio-hydrologique avec deep learning: Les Alpes européennes sont parmi les régions du monde les plus touchées par le changement climatique, avec des taux de recul des glaciers parmi les plus élevés. Les interactions à long terme entre la société, les écosystèmes de montagne et les glaciers de la région soulèvent d'importantes questions sur l'évolution future des glaciers et les impacts environnementaux et socio-économiques qui en découlent. Afin d'évaluer correctement la réponse régionale des glaciers des Alpes françaises au changement climatique, il est nécessaire de disposer d'outils de modélisation adéquats. Dans ce travail, nous explorons de nouvelles façons d'aborder à la fois l'évolution des glaciers et la modélisation glacio-hydrologique à l'échelle régionale. La modélisation de l'évolution des glaciers a traditionnellement été réalisée à l'aide d'approches empiriques ou physiques, dont l'optimisation est de plus en plus difficile compte tenu de la quantité croissante de données disponibles. Ici, nous présentons, à notre connaissance, le premier effort jamais entrepris pour appliquer l'apprentissage profond (i.e. des réseaux neuronaux artificiels profonds) pour simuler l'évolution des glaciers. Comme les système climatique et glaciaire sont tous deux fortement non linéaires, les modèles traditionnels linéaire de bilan de masse offrent une représentation limitée des interactions entre le climat et les glaciers. Nous montrons comment des non-linéarités importantes liées au bilan de masse des glaciers sont capturées par une méthode d'apprentissage profond, ce qui améliore considérablement les performances des modèles par rapport aux méthodes linéaires.Cette nouvelle méthode a été appliquée pour la première fois dans une étude visant à reconstruire les changements annuels du bilan de masse de tous les glaciers des Alpes françaises pour la période 1967-2015. En utilisant des réanalyses climatiques, des données topographiques et des inventaires de glaciers, nous démontrons comment une telle approche peut être utilisée avec succès pour reconstruire les changements de bilan de masse à grande échelle à partir d'observations. Cette étude a également apporté de nouveaux éclairages sur l'évolution des glaciers dans les Alpes françaises au cours du dernier demi-siècle, confirmant les taux de bilan de masse observés plutôt neutres dans les années 1980 et montrant une accélération bien marquée de la perte de masse à partir des années 2000. On constate des différences importantes entre les régions, le massif du Mont-Blanc présentant la perte de masse la plus faible et le Chablais étant le plus touché. Ensuite, nous avons appliqué ce cadre de modélisation pour simuler l'évolution future de tous les glaciers de la région selon de multiples scénarios de changement climatique (N=29). Nos estimations indiquent que la plupart du volume de glace dans la région sera perdue d'ici la fin du XXIe siècle, indépendamment des scénarios climatiques futurs. Nous prévoyons des pertes moyennes de volume des glaciers de 74%, 80% et 88% dans le cadre des scénarios RCP 2.6 (n=3), RCP 4.5 (n=13) et RCP 8.5 (n=13), respectivement. D'ici la fin du XXIe siècle, les Alpes françaises seront en grande partie libres de glace, avec des glaciers ne subsistant que dans les massifs du Mont-Blanc et du Pelvoux.
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
ODINN.jl: A new modular, hybrid, differentiable glacier model
The representation of glacier ice flow dynamics in glacier models in response to climate change, such as basal sliding or rheology, remains a critical challenge, particularly in integrating mechanistic models based on differential equations with the growing availability of observational data. Here, we present ODINN.jl, an open-source, modular framework for hybrid glacier modelling that combines scientific machine learning (SciML) with physical process-based approaches based on partial differential equations (PDEs). The framework is designed to facilitate both forward and inverse simulations of glacier evolution, enabling the assimilation of diverse datasets—such as ice thickness, surface velocities, and climate reanalyses—into a unified modelling ecosystem.We have recently released ODINN.jl v1.0 after almost 5 years of work, providing a new architecture and stable API structured as an interconnected suite of Julia packages, each addressing specific tasks: Sleipnir.jl for data management, Muninn.jl for surface mass balance, Huginn.jl for ice flow dynamics, and ODINN.jl itself as the SciML interface for differentiation, optimisation, and hybrid modelling. This architecture allows users to easily customise model components, swap physical parametrisations, and integrate data-driven models (e.g., neural networks) to represent sub-grid processes or empirical laws. A key feature of this framework is its capacity to leverage automatic differentiation and adjoint methods to optimise model parameters, initial conditions, and statistical regressors. Parallelization is available for both forward simulations and advanced inverse methods, such as universal differential equations (UDEs), to explore poorly understood processes like basal sliding or calving. An early prototype of the model showed its potential to learn hidden laws in a noisy synthetic dataset, and with this new stable release we are now moving to large-scale applications using regional remote sensing and field observations such as high-resolution ice surface velocities and ice thickness. ODINN.jl is compatible with the Open Global Glacier Model (OGGM) ecosystem, enabling simulations for virtually any glacier worldwide using preprocessed datasets (e.g., RGI outlines, DEMs, climate reanalyses). This new modelling framework offers a reproducible, open-source solution to bridge the gap between physical understanding and data-driven discovery. Through modularity, scalability, and open-source collaborative approaches, ODINN.jl aims to explore both methodological advancements and large-scale applied modelling in glaciology.
- 14 March 2026
- Published by Copernicus GmbH
Open AccessODINN.jl: A new modular, hybrid, differentiable glacier model
The representation of glacier ice flow dynamics in glacier models in response to climate change, such as basal sliding or rheology, remains a critical challenge, particularly in integrating mechanistic models based on differential equations with the growing availability of observational data. Here, we present ODINN.jl, an open-source, modular framework for hybrid glacier modelling that combines scientific machine learning (SciML) with physical process-based approaches based on partial differential equations (PDEs). The framework is designed to facilitate both forward and inverse simulations of glacier evolution, enabling the assimilation of diverse datasets—such as ice thickness, surface velocities, and climate reanalyses—into a unified modelling ecosystem.We have recently released ODINN.jl v1.0 after almost 5 years of work, providing a new architecture and stable API structured as an interconnected suite of Julia packages, each addressing specific tasks: Sleipnir.jl for data management, Muninn.jl for surface mass balance, Huginn.jl for ice flow dynamics, and ODINN.jl itself as the SciML interface for differentiation, optimisation, and hybrid modelling. This architecture allows users to easily customise model components, swap physical parametrisations, and integrate data-driven models (e.g., neural networks) to represent sub-grid processes or empirical laws. A key feature of this framework is its capacity to leverage automatic differentiation and adjoint methods to optimise model parameters, initial conditions, and statistical regressors. Parallelization is available for both forward simulations and advanced inverse methods, such as universal differential equations (UDEs), to explore poorly understood processes like basal sliding or calving. An early prototype of the model showed its potential to learn hidden laws in a noisy synthetic dataset, and with this new stable release we are now moving to large-scale applications using regional remote sensing and field observations such as high-resolution ice surface velocities and ice thickness. ODINN.jl is compatible with the Open Global Glacier Model (OGGM) ecosystem, enabling simulations for virtually any glacier worldwide using preprocessed datasets (e.g., RGI outlines, DEMs, climate reanalyses). This new modelling framework offers a reproducible, open-source solution to bridge the gap between physical understanding and data-driven discovery. Through modularity, scalability, and open-source collaborative approaches, ODINN.jl aims to explore both methodological advancements and large-scale applied modelling in glaciology.
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
One mass balance model to rule them all: joint assimilation of remote sensing and glaciological data into MassBalanceMachine
Glacier surface mass balance observations are highly heterogeneous, with point glaciological measurements spanning decadal timescales being limited to a few well-monitored regions, while geodetic observations are available globally with a multi-annual baseline and almost full glacier coverage. Existing surface mass balance modelling approaches can only calibrate their parameters for individual glaciers or regions with available observations. This often implies that a big part of the available observations per glacier (generally glaciological data) cannot be exploited for calibration since they play the role of independent data for validation. There is a need to move towards flexible surface mass balance models, capable of leveraging in a coherent fashion both glaciological and remote sensing data.To address this challenge, we introduce a new version of MassBalanceMachine, a neural network-based model that predicts monthly surface mass balance using topographical features and monthly climate forcing as inputs. The pointwise nature of the model, combined with the global availability of input features thanks to OGGM and ERA5, enables the implementation of custom loss functions to train the model. This loss function allows the model to align with both glaciological measurements and geodetic mass balance observations over multiple decades.By leveraging the interannual variability and point-wise nature from glaciological data and correcting long term biases with geodetic data, MassBalanceMachine can generate reliable high-resolution monthly mass balance predictions for unmonitored glaciers, and correct existing predictions where previous training strategies are known to fail, e.g. in the Alps. This approach exploits data from data-rich sparse regions to make predictions for other unmonitored regions or glaciers, offering a scalable solution for global glacier mass balance estimation.
- 14 March 2026
- Published by Copernicus GmbH
Open AccessOne mass balance model to rule them all: joint assimilation of remote sensing and glaciological data into MassBalanceMachine
Glacier surface mass balance observations are highly heterogeneous, with point glaciological measurements spanning decadal timescales being limited to a few well-monitored regions, while geodetic observations are available globally with a multi-annual baseline and almost full glacier coverage. Existing surface mass balance modelling approaches can only calibrate their parameters for individual glaciers or regions with available observations. This often implies that a big part of the available observations per glacier (generally glaciological data) cannot be exploited for calibration since they play the role of independent data for validation. There is a need to move towards flexible surface mass balance models, capable of leveraging in a coherent fashion both glaciological and remote sensing data.To address this challenge, we introduce a new version of MassBalanceMachine, a neural network-based model that predicts monthly surface mass balance using topographical features and monthly climate forcing as inputs. The pointwise nature of the model, combined with the global availability of input features thanks to OGGM and ERA5, enables the implementation of custom loss functions to train the model. This loss function allows the model to align with both glaciological measurements and geodetic mass balance observations over multiple decades.By leveraging the interannual variability and point-wise nature from glaciological data and correcting long term biases with geodetic data, MassBalanceMachine can generate reliable high-resolution monthly mass balance predictions for unmonitored glaciers, and correct existing predictions where previous training strategies are known to fail, e.g. in the Alps. This approach exploits data from data-rich sparse regions to make predictions for other unmonitored regions or glaciers, offering a scalable solution for global glacier mass balance estimation.
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
Glacier mass balance modeling using a Long Short-Term Memory network
Glacier mass balance is a key indicator of climate change and a central driver of glacier evolution, yet most glaciers worldwide lack long-term in-situ measurements. For estimating glacier mass balance, data-driven models offer a complementary pathway to traditional numerical approaches by learning empirical relationships between climate forcing, topography, and mass balance directly from observations. Here, we develop a recurrent neural network based on a Long Short-Term Memory (LSTM) architecture within the Mass Balance Machine (MBM) framework and evaluate its ability to predict seasonal and annual point surface mass balance across the Swiss Alps. MBM is trained on 30'000 point observations from 30 glaciers and tested on eight glaciers excluded from training to assess spatial generalization. MBM predicts both winter and annual mass balance with high accuracy on unseen glaciers, and its recurrent structure provides a clear advantage by allowing the model to learn temporal dependencies, which improves the representation of seasons with strong accumulation or ablation. Beyond point predictions, MBM produces spatially distributed mass balance maps that closely resemble reference products directly derived from in-situ data, capture elevation-dependent gradients, and yield glacier-wide mass changes consistent with the differencing of repeated terrain models. Monthly outputs further show that the model captures the seasonal transition from winter accumulation to summer ablation and its dependence on elevation with realistic timing and magnitude. These results indicate that a recurrent neural network approach can recover key characteristics of glacier mass balance dynamics from sparse in-situ observations and that the learned relationships are transferable across glaciers with distinct meteorological and topographic settings. The demonstrated generalization skill and suitability for transfer learning highlight the potential of MBM for predicting glacier mass balance in regions with limited or no direct measurements.
- 14 March 2026
- Published by Copernicus GmbH
Open AccessGlacier mass balance modeling using a Long Short-Term Memory network
Glacier mass balance is a key indicator of climate change and a central driver of glacier evolution, yet most glaciers worldwide lack long-term in-situ measurements. For estimating glacier mass balance, data-driven models offer a complementary pathway to traditional numerical approaches by learning empirical relationships between climate forcing, topography, and mass balance directly from observations. Here, we develop a recurrent neural network based on a Long Short-Term Memory (LSTM) architecture within the Mass Balance Machine (MBM) framework and evaluate its ability to predict seasonal and annual point surface mass balance across the Swiss Alps. MBM is trained on 30'000 point observations from 30 glaciers and tested on eight glaciers excluded from training to assess spatial generalization. MBM predicts both winter and annual mass balance with high accuracy on unseen glaciers, and its recurrent structure provides a clear advantage by allowing the model to learn temporal dependencies, which improves the representation of seasons with strong accumulation or ablation. Beyond point predictions, MBM produces spatially distributed mass balance maps that closely resemble reference products directly derived from in-situ data, capture elevation-dependent gradients, and yield glacier-wide mass changes consistent with the differencing of repeated terrain models. Monthly outputs further show that the model captures the seasonal transition from winter accumulation to summer ablation and its dependence on elevation with realistic timing and magnitude. These results indicate that a recurrent neural network approach can recover key characteristics of glacier mass balance dynamics from sparse in-situ observations and that the learned relationships are transferable across glaciers with distinct meteorological and topographic settings. The demonstrated generalization skill and suitability for transfer learning highlight the potential of MBM for predicting glacier mass balance in regions with limited or no direct measurements.
- Open Access
- 13 March 2026
- Published by Copernicus GmbH
Reimagining how Andean glaciers buffered past streamflow droughts
Climate change intensifies water scarcity by increasing the frequency of streamflow droughts. Glaciers play a key role in moderating these events by regulating runoff, but their ongoing retreat threatens this natural resilience. Despite case-specific advances, the regional role of Andean glaciers in shaping streamflow droughts across complex climates and landscapes remains highly uncertain. To address this gap, we use a hybrid glacio-hydrological model that combines process-based glacier mass-balance and ice-flow dynamics with a data-driven runoff representation, allowing us to capture both long-term glacier evolution and short-term hydrological responses. This model is applied to a newly developed dataset of 257 glacierized catchments spanning the Andes (“AndeanGC”; 5–56ºS), which consolidates harmonized hydrological observations, remotely sensed glacier characteristics, and gridded meteorological forcing. The hypothetical future glacier extents correspond to projections under three warming storylines that represent plausible global outcomes: a Paris-aligned pathway limiting warming to 1.5 °C, a current-policy trajectory leading to approximately 2.8 °C of warming, and a high-emission pathway reaching about 4.0 °C. We find that glaciers historically provided substantial buffering of streamflow droughts, but this effect diminishes as glaciers shrink. If past droughts had occurred under the smaller glacier extents projected for the late 21st century under current climate policies, their severity and spatial extent would have increased substantially. Consequently, regional water stress would have intensified markedly. These hypothetical scenarios reveal the previously unquantified regional influence of glaciers on past droughts and illustrate the broader consequences of their decline for water resources. They also highlight the critical need to communicate these changes effectively to support climate-resilient planning and policy.
- 13 March 2026
- Published by Copernicus GmbH
Open AccessReimagining how Andean glaciers buffered past streamflow droughts
Climate change intensifies water scarcity by increasing the frequency of streamflow droughts. Glaciers play a key role in moderating these events by regulating runoff, but their ongoing retreat threatens this natural resilience. Despite case-specific advances, the regional role of Andean glaciers in shaping streamflow droughts across complex climates and landscapes remains highly uncertain. To address this gap, we use a hybrid glacio-hydrological model that combines process-based glacier mass-balance and ice-flow dynamics with a data-driven runoff representation, allowing us to capture both long-term glacier evolution and short-term hydrological responses. This model is applied to a newly developed dataset of 257 glacierized catchments spanning the Andes (“AndeanGC”; 5–56ºS), which consolidates harmonized hydrological observations, remotely sensed glacier characteristics, and gridded meteorological forcing. The hypothetical future glacier extents correspond to projections under three warming storylines that represent plausible global outcomes: a Paris-aligned pathway limiting warming to 1.5 °C, a current-policy trajectory leading to approximately 2.8 °C of warming, and a high-emission pathway reaching about 4.0 °C. We find that glaciers historically provided substantial buffering of streamflow droughts, but this effect diminishes as glaciers shrink. If past droughts had occurred under the smaller glacier extents projected for the late 21st century under current climate policies, their severity and spatial extent would have increased substantially. Consequently, regional water stress would have intensified markedly. These hypothetical scenarios reveal the previously unquantified regional influence of glaciers on past droughts and illustrate the broader consequences of their decline for water resources. They also highlight the critical need to communicate these changes effectively to support climate-resilient planning and policy.
- Open Access
- 13 January 2026
- Published by Copernicus GmbH in The Cryosphere
- Vol. 20 (1) , 171-182
Brief communication: Sensitivity analysis of peak water to ice thickness and temperature: A case study in the Western Kunlun Mountains of the Tibetan Plateau
This study investigates the sensitivity of peak water in the Western Kunlun Mountains of the Tibetan Plateau. Using the Open Global Glacier Model (OGGM), we analyze how variations in inverted initial ice volume and temperature climate forcing under different Shared Socioeconomic Pathways (SSP) affect peak water timing and magnitude. We compare two global ice thickness datasets, revealing substantial differences in the projected peak water timing and magnitude. The results highlight that smaller initial ice volumes lead to earlier peak water occurrences, particularly under the SSP5-8.5 scenario. Temperature bias also notably influences the peak water timing by delaying its date in the region by roughly 13 years for each bias degree. These findings underscore the importance of accurate ice thickness estimates and climate projections for predicting future water availability and informing water management strategies in glacier-dependent regions.
- 13 January 2026
- Published by Copernicus GmbH in The Cryosphere
- Vol. 20 (1) , 171-182
Open AccessBrief communication: Sensitivity analysis of peak water to ice thickness and temperature: A case study in the Western Kunlun Mountains of the Tibetan Plateau
This study investigates the sensitivity of peak water in the Western Kunlun Mountains of the Tibetan Plateau. Using the Open Global Glacier Model (OGGM), we analyze how variations in inverted initial ice volume and temperature climate forcing under different Shared Socioeconomic Pathways (SSP) affect peak water timing and magnitude. We compare two global ice thickness datasets, revealing substantial differences in the projected peak water timing and magnitude. The results highlight that smaller initial ice volumes lead to earlier peak water occurrences, particularly under the SSP5-8.5 scenario. Temperature bias also notably influences the peak water timing by delaying its date in the region by roughly 13 years for each bias degree. These findings underscore the importance of accurate ice thickness estimates and climate projections for predicting future water availability and informing water management strategies in glacier-dependent regions.