simonpf/gprof_ir

Model Card: GPROF IR

Model Details

• Model Name: GPROF IR
• Developer: Simon Pfreundschuh
• License: MIT
• Model Type: Neural Network for Precipitation Retrieval
• Language: Not applicable
• Framework: PyTorch
• Repository: github.com/simonpf/gprof_ir

Model Description

GPROF IR is a satellite precipitation retrieval for geostationary IR observations.

Inputs

• 11 um brightness temperatures from geostationary sensors

Outputs

• Surface precipitation estimates

Training Data

• Training Data Source: Satellite-based observations and collocated ground truth precipitation estimates derived from GPM 2BCMB.
• Data Preprocessing: Normalization

Training Procedure

• Optimizer: AdamW
• Loss Function: Quantile regression
• Training Hardware: 2 NVIDIA RTX 6000
• Hyperparameters: Not exhaustively tuned

Performance

• Evaluation Metrics: Bias, Mean Squared Error (MSE), Mean Absolute Error (MAE), Correlation Coefficient
• Benchmark Comparisons: Compared against ground-based radar.

Intended Use

• Primary Use Case: Satellite-based precipitation retrieval for weather and climate applications
• Potential Applications: Hydrology, extreme weather forecasting, climate research
• Usage Recommendations: Performance may vary across different climate regimes

Ethical Considerations

• Bias Mitigation: Extensive validation against independent datasets

Contact

For questions see corresponding author in reference.