Samudra 2 scales autoregressive neural ocean emulators to finer resolutions with architectural tweaks and dynamic loss, raising upper-ocean temperature R² from 0.56 to 0.87 at 1° and recovering mesoscale features.
Olaf Ronneberger, Philipp Fischer, and Thomas Brox
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
years
2026 6representative citing papers
Q-SRDRN multi-quantile network with pinball loss and per-quantile heads detects extreme precipitation events up to 18 times more effectively than deterministic baselines while preserving augmentation benefits for the median.
IMPA-Net improves extreme convective radar nowcasting by incorporating meteorology-aware multi-scale attention and a three-level asymmetric dynamic loss, raising Heidke Skill Score at ≥45 dBZ from 0.049 to 0.143 versus SimVP while preserving spectral energy better than baselines.
Full conditional distribution modeling outperforms direct binary classification for rare threshold exceedances by learning bulk parameters from moderate events.
A MATLAB/ONNX testbed integrates the Pangu AI model with PID closed-loop control to perform single-input single-output perturbation-response experiments on typhoon track and intensity.
A multimodal GNN ablation for Nordic precipitation nowcasting shows sparse point observations improve station and onset scores while NWP and CRPS losses improve radar-grid performance, indicating local and field skills are distinct targets.
citing papers explorer
-
Samudra 2: Scaling Ocean Emulators across Resolutions
Samudra 2 scales autoregressive neural ocean emulators to finer resolutions with architectural tweaks and dynamic loss, raising upper-ocean temperature R² from 0.56 to 0.87 at 1° and recovering mesoscale features.
-
Multi-Quantile Regression for Extreme Precipitation Downscaling
Q-SRDRN multi-quantile network with pinball loss and per-quantile heads detects extreme precipitation events up to 18 times more effectively than deterministic baselines while preserving augmentation benefits for the median.
-
IMPA-Net: Meteorology-Aware Multi-Scale Attention and Dynamic Loss for Extreme Convective Radar Nowcasting
IMPA-Net improves extreme convective radar nowcasting by incorporating meteorology-aware multi-scale attention and a three-level asymmetric dynamic loss, raising Heidke Skill Score at ≥45 dBZ from 0.049 to 0.143 versus SimVP while preserving spectral energy better than baselines.
-
Forecasting threshold exceedance of atmospheric variables at a specific location
Full conditional distribution modeling outperforms direct binary classification for rare threshold exceedances by learning bulk parameters from moderate events.
-
A Simulation Methodology Testbed for Typhoon Sensitivity Analysis: Framework Development and Perturbation-Response Experiments with the Pangu Weather Model
A MATLAB/ONNX testbed integrates the Pangu AI model with PID closed-loop control to perform single-input single-output perturbation-response experiments on typhoon track and intensity.
-
Pointwise is Pointless? A Multimodal Ablation Study for Precipitation Nowcasting with Graph Neural Networks
A multimodal GNN ablation for Nordic precipitation nowcasting shows sparse point observations improve station and onset scores while NWP and CRPS losses improve radar-grid performance, indicating local and field skills are distinct targets.