CMIP6 bias-adjusted (0.5 deg)

In [ ]:

# Uncomment to install climdata in Google Colab or other environments
# !pip install climdata

# Uncomment to install climdata in Google Colab or other environments # !pip install climdata

Example 1: Basic Usage - Point Extraction¶

Extract climate projection data for a specific location.

In [ ]:

from climdata import ClimData
from climdata.datasets.CMIP_W5E5 import CMIPW5E5

model = 'gfdl-esm4'

for scen in ['ssp126', 'ssp370']:

    # Configure for Berlin, Germany
    overrides = [
        "dataset=cmip_w5e5",                  # Select CMIP-W5E5 dataset
        "lat=52.52",                          # Berlin latitude
        "lon=13.40",                          # Berlin longitude
        f"experiment_id={scen}",              # Climate scenario
        f"source_id={model}",                 # CMIP6 model
        "time_range.start_date=2015-01-01",   # Future period
        "time_range.end_date=2015-12-31",
        "variables=[tas,tasmin,tasmax,pr,rsds,hurs,sfcWind]",
        "data_dir=./data",                    # Local directory for downloaded files
    ]

    extractor = ClimData(overrides=overrides)

    # Initialize CMIP-W5E5 dataset
    cmip_w5e5 = CMIPW5E5(extractor.cfg)

    # Fetch data from ISIMIP repository
    cmip_w5e5.fetch()

from climdata import ClimData from climdata.datasets.CMIP_W5E5 import CMIPW5E5 model = 'gfdl-esm4' for scen in ['ssp126', 'ssp370']: # Configure for Berlin, Germany overrides = [ "dataset=cmip_w5e5", # Select CMIP-W5E5 dataset "lat=52.52", # Berlin latitude "lon=13.40", # Berlin longitude f"experiment_id={scen}", # Climate scenario f"source_id={model}", # CMIP6 model "time_range.start_date=2015-01-01", # Future period "time_range.end_date=2015-12-31", "variables=[tas,tasmin,tasmax,pr,rsds,hurs,sfcWind]", "data_dir=./data", # Local directory for downloaded files ] extractor = ClimData(overrides=overrides) # Initialize CMIP-W5E5 dataset cmip_w5e5 = CMIPW5E5(extractor.cfg) # Fetch data from ISIMIP repository cmip_w5e5.fetch()

Example 2: End-to-End Workflow¶

Chain extraction, imputation, index calculation and NetCDF export in a single call.

In [ ]:

overrides = [
    "dataset=cmip_w5e5",
    "lat=52",
    "lon=13",
    "time_range.start_date=1989-01-01",  # Start date for data extraction
    "time_range.end_date=1989-12-31",    # End date for data extraction
    "variables=[tasmin]",
    "data_dir=./data",                   # Local directory to store raw/intermediate files
    "source_id=GFDL-ESM4",
    "index=tn10p",                       # Climate extreme index to calculate
    "impute=BRITS",                      # Deep-learning gap filling
]

# The workflow runs each action in order
seq = ["extract", "impute", "calc_index", "to_nc"]

extractor = ClimData(overrides=overrides)
result = extractor.run_workflow(actions=seq)

overrides = [ "dataset=cmip_w5e5", "lat=52", "lon=13", "time_range.start_date=1989-01-01", # Start date for data extraction "time_range.end_date=1989-12-31", # End date for data extraction "variables=[tasmin]", "data_dir=./data", # Local directory to store raw/intermediate files "source_id=GFDL-ESM4", "index=tn10p", # Climate extreme index to calculate "impute=BRITS", # Deep-learning gap filling ] # The workflow runs each action in order seq = ["extract", "impute", "calc_index", "to_nc"] extractor = ClimData(overrides=overrides) result = extractor.run_workflow(actions=seq)

Example 3: Explore Available Models and Scenarios¶

Discover what CMIP6 experiments and models are available.

In [ ]:

cmip_w5e5 = CMIPW5E5(extractor.cfg)

# Get available experiment IDs
experiments = cmip_w5e5.get_experiment_ids()
print("Available CMIP6 Experiments:")
for exp in experiments:
    print(f"  - {exp}")

cmip_w5e5 = CMIPW5E5(extractor.cfg) # Get available experiment IDs experiments = cmip_w5e5.get_experiment_ids() print("Available CMIP6 Experiments:") for exp in experiments: print(f" - {exp}")

In [ ]:

# Get available models for a specific experiment
models = cmip_w5e5.get_source_ids(experiment_id='ssp585')
print("\nAvailable Models for SSP5-8.5:")
for model in models:
    print(f"  - {model}")

# Get available models for a specific experiment models = cmip_w5e5.get_source_ids(experiment_id='ssp585') print("\nAvailable Models for SSP5-8.5:") for model in models: print(f" - {model}")

Example 4: Compare Multiple Scenarios¶

Compare low and high emission scenarios for the same location.

In [ ]:

import xarray as xr
import matplotlib.pyplot as plt
from climdata import ClimData
from climdata.datasets.CMIP_W5E5 import CMIPW5E5
# Function to load data for a scenario
def load_scenario(scenario, model='gfdl-esm4'):
    overrides = [
        "dataset=cmip_w5e5",
        "lat=52.52",
        "lon=13.40",
        f"experiment_id={scenario}",
        f"source_id={model}",
        "time_range.start_date=2050-01-01",
        "time_range.end_date=2050-12-31",
        "variables=[tasmin]",
        "data_dir=./data",
    ]
    extractor = ClimData(overrides=overrides)
    cmip = CMIPW5E5(extractor.cfg)
    cmip.fetch()
    cmip.load()
    cmip.extract(point=(extractor.cfg.lon, extractor.cfg.lat))
    return cmip.ds

# Load low and high emission scenarios
ds_low = load_scenario('ssp126')  # Low emissions
ds_high = load_scenario('ssp585') # High emissions

# Plot comparison
plt.figure(figsize=(12, 6))
ds_low['tasmin'].plot(label='SSP1-2.6 (Low emissions)', alpha=0.7)
ds_high['tasmin'].plot(label='SSP5-8.5 (High emissions)', alpha=0.7)
plt.title('Future Temperature Projections for Berlin')
plt.ylabel('Temperature (K)')
plt.xlabel('Year')
plt.legend()
plt.grid(True, alpha=0.3)
plt.tight_layout()
plt.show()

import xarray as xr import matplotlib.pyplot as plt from climdata import ClimData from climdata.datasets.CMIP_W5E5 import CMIPW5E5 # Function to load data for a scenario def load_scenario(scenario, model='gfdl-esm4'): overrides = [ "dataset=cmip_w5e5", "lat=52.52", "lon=13.40", f"experiment_id={scenario}", f"source_id={model}", "time_range.start_date=2050-01-01", "time_range.end_date=2050-12-31", "variables=[tasmin]", "data_dir=./data", ] extractor = ClimData(overrides=overrides) cmip = CMIPW5E5(extractor.cfg) cmip.fetch() cmip.load() cmip.extract(point=(extractor.cfg.lon, extractor.cfg.lat)) return cmip.ds # Load low and high emission scenarios ds_low = load_scenario('ssp126') # Low emissions ds_high = load_scenario('ssp585') # High emissions # Plot comparison plt.figure(figsize=(12, 6)) ds_low['tasmin'].plot(label='SSP1-2.6 (Low emissions)', alpha=0.7) ds_high['tasmin'].plot(label='SSP5-8.5 (High emissions)', alpha=0.7) plt.title('Future Temperature Projections for Berlin') plt.ylabel('Temperature (K)') plt.xlabel('Year') plt.legend() plt.grid(True, alpha=0.3) plt.tight_layout() plt.show()

Example 5: Regional Extraction with Bounding Box¶

Extract data for a region instead of a single point.

In [ ]:

from climdata.datasets.CMIP_W5E5 import CMIPW5E5
from climdata import ClimData

# Configure for Central Europe region
overrides = [
    "dataset=cmip_w5e5",
    "experiment_id=ssp126",              # Low emissions scenario
    "source_id=gfdl-esm4",               # NOAA GFDL Earth System Model
    "time_range.start_date=2050-01-01",
    "time_range.end_date=2050-12-31",
    "variables=[tas,pr]",
    "data_dir=./data",
]

extractor = ClimData(overrides=overrides)
cmip_w5e5 = CMIPW5E5(extractor.cfg)

# Fetch and load
cmip_w5e5.fetch()
cmip_w5e5.load()

# Extract for Central Europe bounding box
box = {
    'lon_min': 5.0,   # Western boundary
    'lon_max': 15.0,  # Eastern boundary
    'lat_min': 47.0,  # Southern boundary
    'lat_max': 55.0   # Northern boundary
}
cmip_w5e5.extract(box=box)

# View spatial data
cmip_w5e5.ds

from climdata.datasets.CMIP_W5E5 import CMIPW5E5 from climdata import ClimData # Configure for Central Europe region overrides = [ "dataset=cmip_w5e5", "experiment_id=ssp126", # Low emissions scenario "source_id=gfdl-esm4", # NOAA GFDL Earth System Model "time_range.start_date=2050-01-01", "time_range.end_date=2050-12-31", "variables=[tas,pr]", "data_dir=./data", ] extractor = ClimData(overrides=overrides) cmip_w5e5 = CMIPW5E5(extractor.cfg) # Fetch and load cmip_w5e5.fetch() cmip_w5e5.load() # Extract for Central Europe bounding box box = { 'lon_min': 5.0, # Western boundary 'lon_max': 15.0, # Eastern boundary 'lat_min': 47.0, # Southern boundary 'lat_max': 55.0 # Northern boundary } cmip_w5e5.extract(box=box) # View spatial data cmip_w5e5.ds

In [ ]:

# Plot a spatial map for a single day
import hvplot.xarray  # noqa: F401

cmip_w5e5.ds['tas'].isel(time=0).hvplot(coastline=True, geo=True)

# Plot a spatial map for a single day import hvplot.xarray # noqa: F401 cmip_w5e5.ds['tas'].isel(time=0).hvplot(coastline=True, geo=True)

Example 6: Save Data to File¶

Export extracted data to NetCDF or CSV format.

In [ ]:

from climdata.datasets.CMIP_W5E5 import CMIPW5E5
from climdata import ClimData

# Configure and extract
overrides = [
    "dataset=cmip_w5e5",
    "lat=52.52",
    "lon=13.40",
    "experiment_id=ssp370",
    "source_id=mpi-esm1-2-hr",
    "time_range.start_date=2080-01-01",
    "time_range.end_date=2080-12-31",
    "variables=[tas,pr,rsds]",
    "data_dir=./data",
]

extractor = ClimData(overrides=overrides)
cmip_w5e5 = CMIPW5E5(extractor.cfg)
cmip_w5e5.fetch()
cmip_w5e5.load()
cmip_w5e5.extract(point=(extractor.cfg.lon, extractor.cfg.lat))

# Save as NetCDF
cmip_w5e5.save_netcdf('berlin_ssp370_2080.nc')

# Save as CSV
cmip_w5e5.save_csv('berlin_ssp370_2080.csv')

print("✅ Data saved successfully!")

from climdata.datasets.CMIP_W5E5 import CMIPW5E5 from climdata import ClimData # Configure and extract overrides = [ "dataset=cmip_w5e5", "lat=52.52", "lon=13.40", "experiment_id=ssp370", "source_id=mpi-esm1-2-hr", "time_range.start_date=2080-01-01", "time_range.end_date=2080-12-31", "variables=[tas,pr,rsds]", "data_dir=./data", ] extractor = ClimData(overrides=overrides) cmip_w5e5 = CMIPW5E5(extractor.cfg) cmip_w5e5.fetch() cmip_w5e5.load() cmip_w5e5.extract(point=(extractor.cfg.lon, extractor.cfg.lat)) # Save as NetCDF cmip_w5e5.save_netcdf('berlin_ssp370_2080.nc') # Save as CSV cmip_w5e5.save_csv('berlin_ssp370_2080.csv') print("✅ Data saved successfully!")

Example 7: Using the ClimData Workflow¶

Use the high-level ClimData workflow for end-to-end processing.

In [ ]:

from climdata import ClimData

# Configure with overrides
overrides = [
    "dataset=cmip_w5e5",
    "lat=12.89",
    "lon=24.25",
    "experiment_id=ssp585",
    "source_id=gfdl-esm4",
    "time_range.start_date=2050-01-01",
    "time_range.end_date=2050-12-31",
    "variables=[tasmin,tasmax,pr]",
    "data_dir=./data",
]

# Initialize extractor
extractor = ClimData(overrides=overrides)

# Extract data
ds = extractor.extract()

# Convert to DataFrame
df = extractor.to_dataframe()

# Save results
result = extractor.save(format='csv')

print(f"Data extracted and saved to: {result.filename}")
df.head()

from climdata import ClimData # Configure with overrides overrides = [ "dataset=cmip_w5e5", "lat=12.89", "lon=24.25", "experiment_id=ssp585", "source_id=gfdl-esm4", "time_range.start_date=2050-01-01", "time_range.end_date=2050-12-31", "variables=[tasmin,tasmax,pr]", "data_dir=./data", ] # Initialize extractor extractor = ClimData(overrides=overrides) # Extract data ds = extractor.extract() # Convert to DataFrame df = extractor.to_dataframe() # Save results result = extractor.save(format='csv') print(f"Data extracted and saved to: {result.filename}") df.head()

Example 8: Calculate Climate Indices¶

Compute extreme indices like frost days or heat waves from CMIP-W5E5 data.

In [ ]:

from climdata import ClimData

# Configure with climate index
overrides = [
    "dataset=cmip_w5e5",
    "lat=52.52",
    "lon=13.40",
    "experiment_id=ssp585",
    "source_id=gfdl-esm4",
    "time_range.start_date=2050-01-01",
    "time_range.end_date=2079-12-31",  # 30 years for climate indices
    "variables=[tasmin,tasmax,pr]",
    "index=tn10p",  # Cold nights (10th percentile of minimum temperature)
    "data_dir=./data",
]

extractor = ClimData(overrides=overrides)

# Extract base data
ds = extractor.extract()

# Calculate index
index_ds = extractor.calc_index()

# View results
print("Climate index calculated:")
index_ds

from climdata import ClimData # Configure with climate index overrides = [ "dataset=cmip_w5e5", "lat=52.52", "lon=13.40", "experiment_id=ssp585", "source_id=gfdl-esm4", "time_range.start_date=2050-01-01", "time_range.end_date=2079-12-31", # 30 years for climate indices "variables=[tasmin,tasmax,pr]", "index=tn10p", # Cold nights (10th percentile of minimum temperature) "data_dir=./data", ] extractor = ClimData(overrides=overrides) # Extract base data ds = extractor.extract() # Calculate index index_ds = extractor.calc_index() # View results print("Climate index calculated:") index_ds

Key Features¶

Scenarios Available¶

• historical: Historical period (typically 1850-2014)
• ssp126: Low emissions (sustainability pathway)
• ssp245: Middle-of-the-road (moderate emissions)
• ssp370: Medium-high emissions
• ssp585: High emissions (fossil-fueled development)

Models Available in ISIMIP3b¶

• gfdl-esm4: NOAA Geophysical Fluid Dynamics Laboratory
• ipsl-cm6a-lr: Institut Pierre-Simon Laplace
• mpi-esm1-2-hr: Max Planck Institute for Meteorology
• mri-esm2-0: Meteorological Research Institute
• ukesm1-0-ll: UK Earth System Modeling

Variables¶

Standard CMIP6 variables are available:

• tas: Near-surface air temperature (K)
• tasmin: Daily minimum temperature (K)
• tasmax: Daily maximum temperature (K)
• pr: Precipitation (kg m⁻² s⁻¹)
• rsds: Surface downwelling shortwave radiation (W m⁻²)
• hurs: Near-surface relative humidity (%)
• sfcWind: Near-surface wind speed (m s⁻¹)
• ps: Surface air pressure (Pa)

Notes¶

1. Data Size: CMIP6 files can be large. Start with short time periods for testing.
2. ISIMIP Client: Requires isimip-client package for data access.
3. Resolution: Data is provided at 0.5° (~55 km) spatial resolution.
4. Bias Adjustment: ISIMIP3b data is bias-adjusted to W5E5 observations.
5. Attribution: When using this data, cite both CMIP6 and ISIMIP appropriately.

References¶

• ISIMIP: https://www.isimip.org/
• CMIP6: https://www.wcrp-climate.org/wgcm-cmip/wgcm-cmip6
• W5E5: https://doi.org/10.48364/ISIMIP.342217