climate-data-analysis

Climate Data Analysis

Overview

Climate data analysis involves working with large multi-dimensional gridded datasets from reanalyses (ERA5, MERRA-2), climate model outputs (CMIP6), observational datasets (GPCC, GHCN), and satellite products. This skill covers the complete workflow from data access to publication-quality visualization.

When to Use This Skill

• Loading and processing NetCDF or GRIB climate datasets
• Computing climatologies, anomalies, and trends
• Spatial and temporal aggregation of climate variables
• Creating climate maps with proper projections (cartopy)
• Accessing CMIP6 model outputs and ERA5 reanalysis data
• Bias correction of climate model outputs
• Calculating climate indices (ENSO, NAO, drought indices)
• Downscaling or regridding climate data

Quick Start

Loading Climate Data with xarray

import xarray as xr
import numpy as np

# Load a NetCDF file
ds = xr.open_dataset("era5_temperature_2020.nc")
print(ds)

# Select a variable and time slice
t2m = ds["t2m"]  # 2m temperature in Kelvin
t2m_celsius = t2m - 273.15  # Convert to Celsius

# Compute annual mean
annual_mean = t2m_celsius.groupby("time.year").mean("time")

# Spatial subset (Europe)
europe = t2m_celsius.sel(
    latitude=slice(75, 35),
    longitude=slice(-15, 45)
)
print(f"Shape: {europe.shape}")
print(f"Time range: {europe.time.values[0]} to {europe.time.values[-1]}")

Climate Climatology and Anomalies

import xarray as xr
import numpy as np

ds = xr.open_dataset("monthly_temperature.nc")
temp = ds["temperature"]

# Compute 30-year climatology (1991-2020 standard)
clim = temp.sel(time=slice("1991", "2020")).groupby("time.month").mean("time")

# Compute anomalies
anomalies = temp.groupby("time.month") - clim

# Rolling trend (10-year)
trend = anomalies.rolling(time=120, center=True).mean()

print(f"Climatology shape: {clim.shape}")
print(f"Anomaly mean: {float(anomalies.mean()):.4f}")

ERA5 Data Access via CDS API

import cdsapi

client = cdsapi.Client()

# Download ERA5 monthly mean 2m temperature
client.retrieve(
    "reanalysis-era5-single-levels-monthly-means",
    {
        "product_type": "monthly_averaged_reanalysis",
        "variable": ["2m_temperature", "total_precipitation"],
        "year": [str(y) for y in range(2000, 2024)],
        "month": [f"{m:02d}" for m in range(1, 13)],
        "time": "00:00",
        "format": "netcdf",
        "area": [90, -180, -90, 180],  # Global
    },
    "era5_monthly_2000_2023.nc",
)