camelsch

CLI tool for downloading, exploring, and extracting data from the CAMELS-CH hydrological dataset (331 Swiss basins, 1981-2020).

Maintenance Status

🟢 Active – Maintained by hydrosolutions

---

Installation

pip install camelsch

This also works inside conda/mamba environments — just activate yours first.

If you use uv for project management:

uv add camelsch

Workflow

camelsch follows a download-then-query workflow:

1. Download the CAMELS-CH dataset once from Zenodo (~1.5 GB).
2. Explore the dataset — list basins, check available variables, view summaries.
3. Extract the data you need — filter by basin, variable, and date range, then export to CSV or Parquet.

All commands read from a local data directory (default ./data/CAMELS_CH). You can change it with --data-dir or the CAMELSCH_DATA_DIR environment variable.

Quick start

# Step 1: Download the dataset from Zenodo (~1.5 GB)
camelsch download

# Step 2: Explore — show dataset summary
camelsch info

# List all basin IDs
camelsch basins
camelsch basins --format json
camelsch basins --format csv --attrs area,p_mean

# Step 3: Extract data
# Static catchment attributes
camelsch attributes --basins 2004,2007 --output attrs.csv

# Time series with filtering
camelsch timeseries --basins 2004 --vars precipitation,discharge_spec \
    --start 1990-01-01 --end 2000-12-31 --output ts.csv

# Batch export to Parquet (with optional attribute merge)
camelsch export --include-attrs --output camels_ch.parquet

Python API

import camelsch

# Load static attributes for specific basins
attrs = camelsch.load_attributes("./data/CAMELS_CH", basin_ids=["2004", "2007"])

# Load time series with filtering
data = camelsch.load_timeseries(
    "./data/CAMELS_CH",
    basin_ids=["2004"],
    variables=["precipitation", "discharge_spec"],
    start_date="1990-01-01",
    end_date="2000-12-31",
)

# List available basins and variables
basins = camelsch.list_basins("./data/CAMELS_CH")
variables = camelsch.list_variables("./data/CAMELS_CH")

Data reference

CAMELS-CH provides daily hydrometeorological time series and static catchment attributes for 331 Swiss basins (1981-2020). The dataset contains 21 time series variables:

Observation-based (9 variables)

Variable	Description
discharge_vol	Discharge volume (m³/s)
discharge_spec	Specific discharge (mm/d)
waterlevel	Water level (m)
precipitation	Basin-mean daily precipitation (mm/d)
temperature_min	Minimum daily temperature (°C)
temperature_mean	Mean daily temperature (°C)
temperature_max	Maximum daily temperature (°C)
rel_sun_dur	Relative sunshine duration (%)
swe	Snow water equivalent (mm)

Simulation-based (12 variables)

Variable	Description
discharge_vol_sim	Simulated discharge volume (m³/s)
discharge_spec_sim	Simulated specific discharge (mm/d)
precipitation_sim	Simulated precipitation (mm/d)
temperature_sim	Simulated temperature (°C)
radiation_sim	Simulated radiation (W/m²)
sun_duration_sim	Simulated sunshine duration (h)
wind_sim	Simulated wind speed (m/s)
rel_humidity_sim	Simulated relative humidity (%)
pet_sim	Simulated potential evapotranspiration (mm/d)
et_sim	Simulated actual evapotranspiration (mm/d)
intercept_et_sim	Simulated interception evapotranspiration (mm/d)
intercept_storage_sim	Simulated interception storage (mm)

Convenience aliases: pet → pet_sim, et → et_sim.

Static attributes cover topography (area, elev_mean), climate (p_mean, t_mean), land cover, soil, and geology.

Full dataset: CAMELS-CH on Zenodo Reference paper: Hoege et al. (2023), CAMELS-CH: hydro-meteorological time series and landscape attributes for 331 catchments in hydrologic Switzerland, Earth Syst. Sci. Data.

For developers

# Install dependencies
uv sync

# Run tests
uv run pytest

# Lint and format
uv run ruff check src/ tests/
uv run ruff format src/ tests/

# Type check
uv run mypy src/

License

MIT