Jikun Liu

Melt ponds function as a decisive regulator of the Arctic ice-albedo feedback and surface energy partition during the summer ablation season. However, their seasonal evolution poses a challenge for parameterization in climate models due to the sub-grid-scale hydro-thermodynamic processes involved in meltwater redistribution. While contemporary data-driven architectures have demonstrated proficiency in instantaneous melt pond mapping, their performance during multi-step temporal roll-out is often compromised by accumulative bias and physical inconsistency, manifesting as systematic geometric degradation and boundary blurring. To address these limitations, this study proposes a Dynamic Physics-Consistent Prediction Model (DPCPM), which reformulates melt pond evolution as a step-wise, physics-constrained temporal propagation problem. The core innovation lies in the dynamic integration of three process-based constraints at each propagation step: thermodynamic energy availability based on surface radiative forcing, morphology-constrained spatial propagation utilizing high-resolution micro-topography, and hydrologic water-balance stability. By enforcing physical consistency within the neural propagation cycle, the DPCPM effectively suppresses the divergent drift inherent in purely stochastic models. The framework was evaluated using multi-source observations including Sentinel-2 imagery, high-resolution digital elevation models, and ERA5 atmospheric forcing collected during the 2020 Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition. Results across the 10 selected melt-pond regions during the 2020 MOSAiC melt period indicate that the DPCPM achieves a pixel-wise Root Mean Square Error (RMSE) between 0.06 and 0.17 for propagation horizons up to one month. The model maintains a high geometric fidelity for ponded regions, achieving Intersection over Union (IoU) values up to 0.53, outperforming unconstrained data-driven baselines. By explicitly bridging sub-grid meltwater dynamics with physical energy constraints, this work provides an observation-constrained and physically consistent framework that may support the development and evaluation of melt pond parameterization in large-scale climate simulations.

Nearshore bathymetry plays a vital role in marine ecological monitoring, coastal zone management, and nautical chart production. Satellite remote sensing has become a primary approach for bathymetry due to its broad spatial coverage, data abundance, and cost-effectiveness. However, single-temporal methods are often limited by transient noise, such as clouds, waves, and ship wakes, which reduce the accuracy. In response, multi-temporal methods have gained increasing attention. By fusing multi-temporal satellite images, these methods can suppress environmental noises, thereby improving the accuracy and stability of depth retrieval. Nonetheless, most existing multi-temporal methods lack physical constraint mechanisms, making them vulnerable to the influence of anomalous images during fusion, and they often overlook temporal variations in water properties and seafloor substrates. These limitations hinder their stability in complex nearshore environments. To address these challenges, this study proposes a novel Multi-Temporal Spaceborne Bathymetry (MTSB) method that integrates ICESat-2 laser altimetry data with multi-temporal Sentinel-2 images. The proposed method uses laser-derived depth points as physical constraints to guide high-quality image fusion. It then applies a stratified modelling strategy based on column laying and sediment classification to adapt to varying conditions. Experiments were conducted in three representative nearshore regions: Culebra, Puerto Rico; Oahu and Niihau, Hawaii. Results show that the MTSB method outperforms existing methods in accuracy. Specifically, the Root Mean Square Error (RMSE) was 1.32 m in Culebra, 1.48 m in Oahu, and 1.61 m in Niihau. Moreover, MTSB showed superior adaptability to environmental disturbances. This research provides a new perspective for high-precision satellite-derived bathymetry and demonstrates promising potential for filling bathymetric data gaps in remote reefs and other inaccessible marine areas.