precipitation-nowcasting

pytorch implemention of trajGRU.

PyTorch-Code for the Paper "SmaAt-UNet: Precipitation Nowcasting using a Small, Attentive UNet-Architecture"

A list of Precipitation Nowcasting papers and related resources.

The repository for paper AA-TransUNet: Attention Augmented TransUNet For Nowcasting Tasks.

An implementation of Deep Generative Model of Radars from DeepMind in PyTorch

This is a Pytorch implementation of PFST-LSTM, a recurrent model for precipitation nowcasting (radar echo extrapolation) as described in the following paper: PFST-LSTM: a SpatioTemporal LSTM Model with Pseudo-flow Prediction for Precipitation Nowcasting

[ICLR 2026] Official Code of: "FlowCast: Advancing Precipitation Nowcasting with Conditional Flow Matching"

Source code for the manuscript "DEUCE v1.0: A neural network for probabilistic precipitation nowcasting with aleatoric and epistemic uncertainties" by Harnist et al. (2024)

Implemented state-of-the-art ConvLSTM and TrajGRU models for precipitation nowcasting, leveraging satellite data with higher spatial (4 km x 4 km) and temporal resolution (15 min) . Created datasets of 1,560 and 779 sequences to train and test. Achieved RMSE of 3.69 mm/hr with 2-hour lead time and RMSE of 8.06 mm/hr with 4-hour lead-time

Code for the paper "MAD-SmaAt-GNet: A multimodal advection-directed network for precipitation nowcasting".

Code for improving precipitation nowcasting models using weighted average ensemble methods.

A ConvLSTM3D model for nowcasting of precipitation using C-Band DWR TERLS.

The M.Sc. Thesis of Bent Harnist (2022) at Aalto University, "Probabilistic Precipitation Nowcasting Using Bayesian Convolutional Neural Networks"

In this project we worked on developing an understanding of RNN and LSTM machine learning paradigms and their application in precipitation nowcasting

华南雷达回波短临预报研究：基于 PyTorch 的 ConvLSTM/LSTM-CNN 与 Vision Transformer 时空序列预测，含数据清洗、训练和可视化评估 | South China radar echo nowcasting with PyTorch, ConvLSTM and ViT, including radar data preprocessing, model training and forecast visualization.

Curated machine learning papers, datasets and tools for precipitation nowcasting, retrieval and downscaling

Solafune's Precipitation Nowcasting From Space competition: estimating GPM-IMERG rainfall maps from the last 30 minutes of Himawari, GOES and Meteosat imagery. One shared ConvNeXt U-Net with a rain/no-rain hurdle head, validated on unseen locations.

[ACCV26] Physics-Guided Flow-Map Matching for Precipitation Nowcasting

Expanded list of precipitation nowcasting papers, models, etc.

SALT: Symmetry-Augmented Level-Set Training for Radar Echo Extrapolation (code coming upon acceptance)