Tancrède Pierre Marie Leger

Current glacier model-data comparison methods often lack quantitative, automated validation for reconstructing complex palaeoglaciers and palaeo ice fields over millennia. Rigorous validation and automatic evaluation is critical for improving the reliability of glacier reconstructions and their use in understanding past climate and future projections. We introduce an open-source, Python-based comparison tool (GLAMODAT version 1.0) that automatically evaluates palaeoglacier model simulations against dated constraints on glacier history. GLAMODAT successfully quantifies model-data mismatches and demonstrates robustness in both synthetic and real-world case studies. It effectively tracks multiple phases of advance and retreat periods within mountain valleys. It is designed for validating single palaeoglacier simulations or entire ensembles by integrating weighted misfit scoring and adaptive parameter space exploration. It enables advancements in palaeoglacier reconstructions, climate reconstructions, and parameter studies in glacier modelling.

Mountain glacier evolution since the last interglacial remains poorly constrained, with limited spatial and temporal coverage. Conventional modelling approaches typically target major climatic events, operate at coarse spatial resolution over limited spatial domains, or employ simplified representations of ice dynamics. Here, we address these limitations by applying the Instructed Glacier Model (IGM), to reconstruct, for the first time, mountain glacier evolution since ∼130 ka at 500 m resolution across nine mountain ranges in North America, South America, Eurasia, and Africa. We perform 707 parameter-calibration simulations by varying paleoclimate and ice-dynamic parameters, and validate model performance by assessing glacier extent and ice thickness. The model outputs are evaluated using a spatial frequency map approach, which identifies a set of acceptable model results rather than a single best-fit simulation. As a result, we identify areas of robust agreement and those sensitive to parameter choices, providing a systematic way to visualise spatial uncertainty and glacier–climate-topography interactions. Together, our framework is a scalable foundation for next-generation, uncertainty-aware reconstructions turning glacier modelling at orbital-timescales into a reproducible, expandable workflow that can be deployed across mountain ranges worldwide. All data are publicly available at https://archive.sigma2.no/dataset/evolution-of-mountain-glaciers-since-the-last-interglacial (Barndon et al., 2026b) and the animations at https://av.tib.eu/series/2022 (Barndon et al., 2026a).

The AlpIce‐1 database is the first compilation of published geomorphological and geochronological markers constraining build‐up, culmination and decay of glaciers from the Last Glacial Maximum through the Holocene across the entire European Alps. AlpIce‐1 integrates more than 4200 surface‐exposure and radiocarbon ages, homogenized, quality controlled and evaluated for their palaeoglaciological context, compiled from 268 peer‐reviewed publications and over 1300 sites across the European Alps. The database includes ages originally obtained from a wide range of landforms and deposits including moraine ridges, lakes, mires, sediment outcrops, ice cores, bedrock surfaces, caves and archaeological sites. It documents chronological constraints on more than 1400 palaeoglacial events, including past ice‐margin stabilization, ice (re‐)advance, ice retreat, ice‐free or ice‐cover conditions. A GIS catalogue of more than 260 dated landforms (e.g. moraines, rock glaciers, landslide deposits) complements the database. AlpIce‐1 is designed as an open‐access resource supporting both field and modelling studies across the Alps and is intended for all communities interested in the late Quaternary. In light of emerging next‐generation numerical ice models, AlpIce‐1 offers an extensive and robust empirical basis for evaluating future (palaeo)glacier simulations.

Reconstructing the transport histories and provenances of glacial sediments and ice-contact deposits (e.g. tills, moraines) in formerly glaciated regions remains a major challenge, particularly at icefield- to ice-sheet scales and over multi-millennial timescales. Yet such reconstructions are central to key questions in Quaternary science, including estimates of past glacial erosion rates and sediment fluxes, the role of subglacial sediment storage in erosion reduction, or the reconstruction of past ice-flow dynamics, ice divides, and transfluences. While numerical modelling can enable one to reproduce past glacial sediment transport via coupling glacier models with particle tracking, this becomes computationally unfeasible over large spatial domains and paleo timescales using traditional computing. As a result, no study to date has simulated glacial sediment transport using large particle numbers (tens of millions) across continental-scale icefields such as the one occupying the European Alps during the Last Glacial Maximum (LGM). Here, we overcome this limitation using the Instructed Glacier Model (IGM), which allows the coupling of 3D Lagrangian particle tracking with high-resolution glacier simulations, both accelerated on Graphics Processing Units (GPU). This unlocks the modelling of ice advection of millions of particles at minimal additional computational cost, allowing simulations of glacial sediment transport across the European Alps over multi-millennial timescales (40–18 ka) and at the unprecedented spatial resolution of 300 m. We achieve ∼ 50× faster computation tracking 20 million particles across the Alps using a single GPU instead of 60 CPU threads. In doing so, we produce the first Alps-wide modelling reconstruction of glacial sediment transport during the LGM, using process-based particle seeding schemes to represent both subglacial (e.g. abrasion, plucking) and supraglacial (e.g. rockfall, landslides) sediment sourcing. Results are analysed through complementary “sink-to-source” (deposit provenance) and “source-to-sink” (potential depositional pathways) analyses, enabling us to reconstruct the LGM glacial transport of numerous ice-contact deposits and surface lithologies across the Alps. We find that supraglacially sourced glacial sediments are typically eroded earlier, experience longer glacier residence times, and undergo greater cumulative ice-free exposure than those of subglacial origin, with implications for the interpretation of cosmogenic nuclide inheritance in glacial deposits. Our new coupled glacier-particle modelling framework opens avenues for quantitative model-data comparisons using glacial geomorphology and provides a powerful tool for reconstructing paleo ice dynamics, sediment provenance, and Quaternary glacial landscape evolution.

Reconstructing the transport histories and provenances of glacial sediments and ice-contact deposits (e.g. tills, moraines) in formerly glaciated regions remains a major challenge, particularly at icefield- to ice-sheet scales and over multi-millennial timescales. Yet such reconstructions are central to key questions in Quaternary science, including estimates of past glacial erosion rates and sediment fluxes, the role of subglacial sediment storage in erosion buffering, or the reconstruction of past ice-flow dynamics, ice divides, and transfluences. While numerical modelling can enable one to reproduce past glacial sediment transport via coupling glacier models with Lagrangian particle tracking, this becomes computationally unfeasible over large spatial domains and paleo timescales using traditional computing. As a result, no study to date has simulated glacial sediment transport using large particle numbers (tens of millions) across continental-scale icefields such as the European Alps during the Last Glacial Maximum (LGM): a pre-requisite given the ubiquitous nature of sediments in glacier systems. In this study, we overcome this limitation through a new coupling of 3D Lagrangian particle tracking with Graphics-Processing-Units (GPU)-accelerated, high-resolution glacier simulations based on the deep-learning-enhanced Instructed Glacier Model (IGM). Our approach unlocks the ice advection of tens of millions of particles at minimal additional computational cost, allowing simulations of glacial sediment transport across the European Alps over multi-millennial timescales (40–18 ka) and at an unprecedented spatial resolution of 300 m. In doing so, we produce the first Alps-wide modelling reconstruction of glacial sediment transport during the LGM, using process-based particle seeding schemes to represent both subglacial (e.g. abrasion, plucking) and supraglacial (e.g. rockfall, landslides) sediment sourcing. Results are analysed through complementary ‘sink-to-source’ (deposit provenance) and ‘source-to-sink’ (potential depositional pathways) analyses, enabling us to reconstruct the LGM glacial transport of numerous ice-contact deposits and surface lithologies across the Alps. We find that supraglacially sourced glacial sediments are typically eroded earlier, experience longer glacier residence times, and undergo greater cumulative ice-free exposure than those of subglacial origin, with implications for the interpretation of cosmogenic nuclide inheritance in glacial deposits. Our new coupled glacier-particle modelling framework opens avenues for quantitative model-data comparisons using glacial geomorphology and provides a powerful tool for reconstructing paleo ice dynamics, sediment provenance, and Quaternary glacial landscape evolution.

Reconstructing the transport histories and provenances of glacial sediments and ice-contact deposits (e.g. tills, moraines) in formerly glaciated regions remains a major challenge, particularly at icefield- to ice-sheet scales and over multi-millennial timescales. Yet such reconstructions are central to key questions in Quaternary science, including estimates of past glacial erosion rates and sediment fluxes, the role of subglacial sediment storage in erosion buffering, or the reconstruction of past ice-flow dynamics, ice divides, and transfluences. While numerical modelling can enable one to reproduce past glacial sediment transport via coupling glacier models with Lagrangian particle tracking, this becomes computationally unfeasible over large spatial domains and paleo timescales using traditional computing. As a result, no study to date has simulated glacial sediment transport using large particle numbers (tens of millions) across continental-scale icefields such as the European Alps during the Last Glacial Maximum (LGM): a pre-requisite given the ubiquitous nature of sediments in glacier systems. In this study, we overcome this limitation through a new coupling of 3D Lagrangian particle tracking with Graphics-Processing-Units (GPU)-accelerated, high-resolution glacier simulations based on the deep-learning-enhanced Instructed Glacier Model (IGM). Our approach unlocks the ice advection of tens of millions of particles at minimal additional computational cost, allowing simulations of glacial sediment transport across the European Alps over multi-millennial timescales (40–18 ka) and at an unprecedented spatial resolution of 300 m. In doing so, we produce the first Alps-wide modelling reconstruction of glacial sediment transport during the LGM, using process-based particle seeding schemes to represent both subglacial (e.g. abrasion, plucking) and supraglacial (e.g. rockfall, landslides) sediment sourcing. Results are analysed through complementary ‘sink-to-source’ (deposit provenance) and ‘source-to-sink’ (potential depositional pathways) analyses, enabling us to reconstruct the LGM glacial transport of numerous ice-contact deposits and surface lithologies across the Alps. We find that supraglacially sourced glacial sediments are typically eroded earlier, experience longer glacier residence times, and undergo greater cumulative ice-free exposure than those of subglacial origin, with implications for the interpretation of cosmogenic nuclide inheritance in glacial deposits. Our new coupled glacier-particle modelling framework opens avenues for quantitative model-data comparisons using glacial geomorphology and provides a powerful tool for reconstructing paleo ice dynamics, sediment provenance, and Quaternary glacial landscape evolution.

To compensate for the high computational costs when modelling large-scale mountain glaciers, ice fields or ice sheets over multiple millennia, it is common practice to coarsen the spatial resolution of numerical models to 1 km or more, which is not sufficient to describe complex valley topographies. Here, we examine the influence of spatial resolution by modelling a growing and retreating exemplary ice field in the European Alps at resolutions ranging from 50 m to 2 km using the Instructed Glacier Model (IGM). We find that while ice-covered areas remain similar, ice volume increases substantially with coarser resolution. Compared to the reference run at 50 m spatial resolution, model results at a resolution of 300 m and finer are comparable and sufficiently accurate to simulate topographically constrained ice flow. However, at resolutions coarser than ∼800 m, topographic resampling artificially lowers slope angles and mountain peaks, providing a larger accumulation area at high altitudes, with thicker glaciers that are typically warm-based, while thinner glaciers at fine resolutions remain cold-based. Raised valley floors at coarse resolutions result in slower-flowing ice with increased thickness and glacial response times. The resulting hysteresis between climate forcing and glacial response at coarse resolutions is only partially decreased with slower temperature change. Seemingly stable model results at coarse resolutions may be misleading and accurate glacier geometries might arise from parameter choices that compensate for poorly resolved topography. Similar non-linear and altitudinal-dependent resolution effects are likely in mountain regions worldwide and emphasize the need for model advances to enable simulations at sufficiently high spatial resolutions to accurately resolve glacier dynamics.

Reconstructing the transport histories and provenances of glacial sediments and ice-contact deposits (e.g. tills, moraines) in formerly glaciated regions remains a major challenge, particularly at icefield- to ice-sheet scales and over multi-millennial timescales. Yet such reconstructions are central to key questions in Quaternary science, including estimates of past glacial erosion rates and sediment fluxes, the role of subglacial sediment storage in erosion buffering, or the reconstruction of past ice-flow dynamics, ice divides, and transfluences. While numerical modelling can enable one to reproduce past glacial sediment transport via coupling glacier models with Lagrangian particle tracking, this becomes computationally unfeasible over large spatial domains and paleo timescales using traditional computing. As a result, no study to date has simulated glacial sediment transport using large particle numbers (tens of millions) across continental-scale icefields such as the European Alps during the Last Glacial Maximum (LGM): a pre-requisite given the ubiquitous nature of sediments in glacier systems. In this study, we overcome this limitation through a new coupling of 3D Lagrangian particle tracking with Graphics-Processing-Units (GPU)-accelerated, high-resolution glacier simulations based on the deep-learning-enhanced Instructed Glacier Model (IGM). Our approach unlocks the ice advection of tens of millions of particles at minimal additional computational cost, allowing simulations of glacial sediment transport across the European Alps over multi-millennial timescales (40–18 ka) and at an unprecedented spatial resolution of 300 m. In doing so, we produce the first Alps-wide modelling reconstruction of glacial sediment transport during the LGM, using process-based particle seeding schemes to represent both subglacial (e.g. abrasion, plucking) and supraglacial (e.g. rockfall, landslides) sediment sourcing. Results are analysed through complementary ‘sink-to-source’ (deposit provenance) and ‘source-to-sink’ (potential depositional pathways) analyses, enabling us to reconstruct the LGM glacial transport of numerous ice-contact deposits and surface lithologies across the Alps. We find that supraglacially sourced glacial sediments are typically eroded earlier, experience longer glacier residence times, and undergo greater cumulative ice-free exposure than those of subglacial origin, with implications for the interpretation of cosmogenic nuclide inheritance in glacial deposits. Our new coupled glacier-particle modelling framework opens avenues for quantitative model-data comparisons using glacial geomorphology and provides a powerful tool for reconstructing paleo ice dynamics, sediment provenance, and Quaternary glacial landscape evolution.