Version 1.0 - Initial Public Release

Statistical Downscaling of Climate Models for Regional Hydrology

This release marks the first public availability of methods developed during 2006-2015 doctoral and post-doctoral research.

Key Features

• Multi-GCM ensemble downscaling supporting 5 models (CGCM2, CSIRO2, ECHAM4, HadCM3, PCM)
• SRES scenario support (A1, A2, B1, B2)
• SST teleconnection integration (ENSO, IOD)
• Moran's I spatial autocorrelation preservation for stochastic weather generation
• 276 R scripts across 8 analysis modules
• Comprehensive documentation for reproducibility

Components

Core Analysis Modules

• ML-regression-multiGCM: Multi-GCM machine learning regression
• multi-regression model: Multiple regression for climate variables
• Stochastic: Spatial weather generation using Moran's I
• arima: ARIMA time series modeling
• data-processing: Regional data extraction
• filling-NA: Missing data imputation
• cross-correlation: GCM-observation correlation analysis
• basic time series analysis: Exploratory analysis

Capabilities

• Calibration/verification periods: 1971-1985/1986-1999 or 1971-1999/2000-2006
• Predictions: 1971-2099
• Configuration-driven approach for flexible model combinations
• Supports GCMs only, High-resolution only, GCMs+High-res, with/without SSTs

Research Period

2006-2015

Documentation

See README.md for complete methodology, usage instructions, and project overview.

License

MIT with academic use notice

---

🤖 Generated with Claude Code