Abstract. Glacier mass balance is typically estimated using a range of in situ measurements, remote sensing measurements, and physical and temperature index modelling techniques. With improved data collection and access to large datasets, data-driven techniques have recently gained prominence in modelling natural processes. The most common data-driven techniques used today are linear regression models and, to some extent, non-linear machine learning models such as artificial neural networks. However, the entire host of capabilities of machine learning modelling has not been applied to glacier mass balance modelling. This study used monthly meteorological data from ERA5-Land to drive four machine learning models: random forest (ensemble tree type), gradient-boosted regressor (ensemble tree type), support vector machine (kernel type), and artificial neural networks (neural type). We also use ordinary least squares linear regression as a baseline model against which to compare the performance of the machine learning models. Further, we assess the requirement of data for each of the models and the requirement for hyperparameter tuning. Finally, the importance of each meteorological variable in the mass balance estimation for each of the models is estimated using permutation importance. All machine learning models outperform the linear regression model. The neural network model depicted a low bias, suggesting the possibility of enhanced results in the event of biased input data. However, the ensemble tree-based models, random forest and gradient-boosted regressor, outperformed all other models in terms of the evaluation metrics and interpretability of the meteorological variables. The gradient-boosted regression model depicted the best coefficient of determination value of 0.713 and a root mean squared error of 1.071 m w.e. The feature importance values associated with all machine learning models suggested a high importance of meteorological variables associated with ablation. This is in line with predominantly negative mass balance observations. We conclude that machine learning techniques are promising in estimating glacier mass balance and can incorporate information from more significant meteorological variables as opposed to a simplified set of variables used in temperature index models.

Reuyl is an 85‐km‐diameter crater located east of the Aeolis Dorsa region. We present results from geomorphic mapping and high‐resolution image analyses of Reuyl to understand the provenance of alluvial fans, superposed deposits, channels, and sinuous ridges. We distinguish 29 fans or other deposits associated with channels based on their characteristics. Two sinuous ridges are associated with the alluvial fans formed between the wall and peak region. On the southern floor, a sinuous ridge associated with an alluvial fan with channel inversion is orthogonally superposed by another wall‐originated fan deposit. On the western floor, a sinuous ridge associated with an alluvial fan is superposed by wall‐originated deposits. We also observe bajadas, stacked deposits, deposits with incised channels, and overlapping fan ridges, which suggest diverse and long‐lived depositional activity related to a fluvial environment. The central peak is surrounded by a mound deposit with several discontinuous channels. One large fan from the western wall appears more intact than the others, being less eroded and lacking ridges. Reuyl aged as ~3.63 Ga, whereas the first‐order analysis of the fan deposits implies ~3.5 Ga and the large intact deposit is estimated to belong to the Amazonian epoch. We suggest that Reuyl crater likely witnessed and recorded the transition from alluvial fan to large younger intact deposits, which possibly suggests a decrease in water/volatile‐related activity and/or change in sediment supply. Overall, the diverse superposed fan deposits within Reuyl reveal noncoeval activity and provide comprehensive evidence for long‐lived fluvial activity during the post‐Noachian epoch.

Abstract. Glacier mass balance is typically estimated using a range of in-situ measurements, remote sensing measurements, and physical and temperature index modelling techniques. With improved data collection and access to large datasets, data-driven techniques have recently gained prominence in modelling natural processes. The most common data-driven techniques used today are linear regression models and, to some extent, non-linear machine learning models such as artificial neural networks. However, the entire host of capabilities of machine learning modelling has not been applied to glacier mass balance modelling. This study used monthly meteorological data from ERA5-Land to drive four machine learning models: random forest (ensemble tree type), gradient-boosted regressor (ensemble tree type), support vector machine (kernel type) and artificial neural networks (neural type). We also use ordinary least squares linear regression as a baseline model against which to compare the performance of the machine learning models. Further, we assess the requirement of data for each of the models and the requirement for hyperparameter tuning. Finally, the importance of each meteorological variable in the mass balance estimation for each of the models is estimated using permutation importance. All machine learning models outperform the linear regression model. The neural network model depicted a low bias, suggesting the possibility of enhanced results in the event of biased input data. However, the ensemble tree-based models, random forest and gradient-boosted regressor outperformed all other models in terms of the evaluation metrics and interpretability of the meteorological variables. The gradient-boosted regression model depicted the best coefficient of determination value of 0.713. The feature importance values associated with all machine learning models suggested high importance to meteorological variables associated with ablation. This is in line with predominantly negative mass balance observations. We conclude that machine learning techniques are promising in estimating glacier mass balance and can incorporate information from more significant meteorological variables as opposed to a simplified set of variables used in temperature index models.

No abstract