Physics-constrained generative machine learning-based high-resolution downscaling of Greenland’s surface mass balance and surface temperature
Nils Bochow, Philipp Hess, Alexander Robinson · 2025
Read Physics-constrained generative machine learning-based high-resolution downscaling of Greenland’s surface mass balance and surface temperature by Nils Bochow, Philipp Hess, Alexander Robinson (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-22485.txt.
Introduction
Nils Bochow, Philipp Hess, and Alexander Robinson sharpen coarse climate maps of Greenland with a generative model that respects physics. Fine maps of surface mass balance and temperature drive sea-level forecasts, but running models at fine scale is costly. Their 2025 method uses a consistency model, a fast generative tool, trained on output from a regional climate model. It takes wide-grid fields and redraws them at up to thirty-two times the detail, down to five-kilometer spacing, in a few steps.
Physical rules are built into the training so the fine maps stay plausible, not just pretty. Ice-sheet features guide the conditioning. The authors frame the gain as speed with honesty: high detail without weeks of computing. Such maps can feed ice models that project how Greenland melts.
Catalog record
| CREATOR | Nils Bochow, Philipp Hess, Alexander Robinson |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-22485
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Physics-constrained generative machine learning-based high-resolution downscaling of Greenland’s surface mass balance and surface temperature Nils Bochow, Philipp Hess, Alexander Robinson Abstract Accurate, high-resolution projections of the Greenland ice sheet’s surface mass balance (SMB) and surface temperature are essential for understanding future sea-level rise, yet current approaches are either computationally demanding or limited to coarse spatial scales. Here, we introduce a novel physics-constrained generative modeling framework based on a consistency model (CM) to downscale low-resolution SMB and surface temperature fields by a factor of up to 32 (from 160 km to 5 km grid spacing) in a few sampling steps. The CM is trained on monthly outputs of the regional climate model MARv3.12 and conditioned on ice-sheet topography and insolation. By enforcing a hard conservation constraint during inference, we ensure approximate preservation of SMB and temperature sums on the coarse spatial scale as well as robust generalization to extreme climate states without retraining. On the test set, our constrained CM achieves a continued ranked probability score of 6.31 mmWE for the SMB and 0.1 K for the surface temperature, outperforming interpolation-based downscaling. Together with spatial power-spectral analysis, we demonstrate that the CM faithfully reproduces variability across spatial scales. We further apply bias-corrected outputs of the NorESM2 Earth System Model as inputs to our CM, to demonstrate the potential of our model to directly downscale ESM fields. Our approach delivers realistic, high-resolution climate forcing for ice-sheet simulations with fast inference and can be readily integrated into Earth-system and ice-sheet model workflows to improve projections of the future contribution to sea-level rise from Greenland and potentially other ice sheets and glaciers too. Keywords Generative modeling \cdot Greenland Ice Sheet \cdot Surface Mass Balance \cdot Downscaling 1 Introduction The Greenland ice sheet (GrIS) is the second largest ice sheet with an ice volume of 7.4 m sea level equivalent [ 1 ] . Anthropogenic warming has driven a significant acceleration of Greenland mass loss over recent decades [ 2 ] . Currently, the mass loss of the GrIS is approximately evenly driven by dynamic discharge and a decrease in the surface mass balance (SMB), with ca. 60% of GrIS ice loss due to the changes at the surface [ 3 , 4 ] . However, by the end of this century, the importance of surface processes is projected to increase as marine outlets retreat and dynamic processes will play a smaller role [ 5 , 4 , 6 ] . That means that the SMB will be the determining factor of the overall mass balance of the GrIS. Modeling the GrIS surface mass balance demands fine resolution to capture the steep orographic gradients along the ice-sheet margin that govern precipitation patterns [ 7 ] . High-resolution simulations also resolve the narrow ablation zone and peripheral outlet glaciers, where most of the melt and runoff occur [ 8 ] . These high-resolution SMB fields then feed into ice-sheet model intercomparisons (e.g., ISMIP6), improving projections of dynamic thinning and sea-level contribution [ 5 ] . Therefore, it is crucial to have realistic and high-resolution projections of the SMB in a changing climate. Current methods to calculate the SMB have several disadvantages. They are either expensive to run, based on simple parameterisations or have low resolution. On short time scales, i.e., centennial scales, regional climate models such as MAR [ 9 , 10 ] or RACMO [ 11 ] are used to simulate the surface mass balance. These RCMs are specifically designed to simulate the polar regions. However, they are computationally expensive to run and are usually not accounting for a changing ice-sheet topography. On longer time scales, the SMB is usually simulated via simple parameterisation schemes such as the positive degree days method (PDD) [ 12 , 13 , 14 ] or energy balance models [ 15 , 16 , 17 ] . Alternatively, the SMB can be directly derived from ESM runs as a sum of precipitation, melt and runoff in a first approximation [ 18 , 19 ] . However, the ESM-derived SMB fields are limited to the, often coarse, resolution of the ESM and cannot resolve the steep gradients of the ice sheet, leading to biases in SMB estimates [ 20 ] . Given these computational constraints and resolution trade-offs across centennial and longer scales, computationally fast approaches are required to efficiently produce high-resolution SMB estimates. Previous approaches to downscaling the SMB of Greenland are often based on statistical downscaling [ 21 , 22 ] . These approaches leverage the relationship between elevation, SMB and temperature, and have shown promising results in downscaling RCM results to higher resolutions. However, these statistical methods are inherently limited by their prescribed relationships between melt and elevation, which, for example, do not hold for the Antarctic ice sheet and can therefore not easily be extended to other regions [ 22 ] . Additionally, these statistical methods cannot easily integrate complementary inputs that could improve the downscaling task [ 22 ] . Therefore, novel machine-learning techniques, which can be trained to learn complex relationships, including topographic and climatological patterns, from high-resolution simulations and then run orders of magnitude faster are a promising approach. In contrast to previous statistical approaches, we aim to provide a framework in this work that is able to downscale SMB and T_{s} fields from other sources than RCMs, that is, deriving high-resolution fields directly from ESM output or from simpler physics-based models. Machine learning (ML) methods have gained tremendous attention in climate science in the last years, ranging from reconstruction of climate fields [ 23 ] , ML-based global circulation and weather models [ 24 , 25 ] , downscaling [ 26 ] and weather forecasting [ 27 , 28 , 29 ] . In the glaciological community, ML has, for example, been used to improve ice-flow simulations [ 30 , 31 ] , uncover flow laws of Antarctic ice shelves [ 32 ] , improve upon traditional ice-shelf melt parameterisations [ 33 ] , forecast glacier-wide SMB [ 34 , 35 , 36 ] or as emulators for subglacial hydrology models [ 37 ] . Generative modeling approaches, and diffusion models in particular, have shown promising results in downscaling of ESM fields [ 38 , 39 , 40 , 26 ] . The principle behind downscaling with generative diffusion models is based on so-called denoising, where a network is first trained to remove noise added to a high-resolution sample. Once trained, the network can then be applied for downscaling by iteratively generating realistic small-scale patterns from noise while following the large-scale patterns of the low-resolution ESM. However, most applications of generative downscaling so far are (i) limited to univariate fields (e.g. precipitation), (ii) struggle with training instabilities or mode collapse [ 41 ] , (iii) require a large amount of sample steps to generate realistic fields [ 42 ] and/or (iv) lack physical mass conservation [ 26 ] . Here, we introduce a generative modeling-based methodology to derive high-resolution surface mass balance and temperature at the surface ( T_{s} ) fields for the Greenland ice sheet. The SMB and T_{s} are the most important, minimal surface forcing fields needed to drive an ice sheet model. We use a generative consistency model (CM) [