Module 1. Observational Climate Pipeline for Ethiopia and VECTRI Baseline Setup¶

Purpose¶

Build a clean observational baseline using gridded rainfall and temperature datasets, preprocess them consistently, define the Ethiopia target grid, and prepare climate inputs for VECTRI.

Recommended data sources¶

Sessions¶

1.0 Environment setup and project folder structure¶

Full step-by-step guide: Environment setup and folder structure.

• Install VS Code and the Python / Jupyter extensions.
• Create and activate the Python virtual environment (.venv).
• Install all core packages: numpy, pandas, matplotlib, cftime, cf_xarray, openpyxl, shapely, scipy, requests, cartopy, geopandas, rioxarray, rasterio, regionmask, salem, netCDF4.
• Create the canonical folder structure for data, processed outputs, ancillary files, QC reports, and logs.

data/
  raw/chirps/               ← raw yearly CHIRPS NetCDF downloads
  processed/chirps/         ← QC-passed, clipped, merged files
  ancillary/grid_mask/      ← Ethiopia land mask and target-grid definition
reports/
  qc_chirps/                ← diagnostic plots and QC reports
  logs/                     ← download and processing logs

1.1 Observational data access and downloading¶

• Download CHIRPS rainfall for Ethiopia — step-by-step lesson: Download CHIRPS rainfall for Ethiopia.
• Download observational temperature from ERA5-Land or CHIRTS-daily.
• Organize file structure for historical and operational use.

1.2 Daily ingestion, preprocessing, and quality control¶

• Detailed guide: CHIRPS ingestion, preprocessing, and QC.

• Standardize file naming and metadata.

• Check missing files, duplicate days, invalid values, and time continuity.
• Harmonize units and calendars.

Produce diagnostic plots:

• Daily time-series checks
• Annual cycle plots
• Rainfall distribution plots
• Missing-data summaries
• Domain maps for quick inspection

1.3 Ethiopia target domain and target grid definition¶

• Detailed guide: Ethiopia target grid and land mask.

• Define the Ethiopia analysis domain.

• Define the CHIRPS target grid for all downstream processing.
• Build an Ethiopia land mask on the CHIRPS grid.
• Apply the same target grid and mask consistently across all datasets.

1.4 Historical split design¶

• Period splits in practice: Observational standardization and VECTRI prep.

Explain why these periods are used and how they map to model fitting, verification, and live operations.

1.5 Preparing climate inputs for VECTRI¶

• Structure, masking, and versioning: Observational standardization and VECTRI prep and Documentation and reproducibility.

• Convert rainfall and temperature into VECTRI-ready format.

• Align spatial grid, time step, units, and file structure.
• Run VECTRI using the preprocessed observational inputs.
• Check VECTRI output consistency before moving to forecast-driven runs.

Expected outputs¶

• Clean Ethiopia rainfall and temperature baseline dataset
• Ethiopia CHIRPS target grid and land mask
• QC and diagnostic plots
• Historical split ready for calibration, validation, and operations
• Baseline VECTRI simulation driven by observed climate