climate-modelling-python

Python-based climate analysis covering temperature trends and dynamic carbon cycle modelling.

---

1. Canada Climate Analysis

Notebook: canada_climate_analysis.ipynb Data: canada-TAVG-monthly.txt (Berkeley Earth), co2_annmean_gl.txt (NOAA GML)

What it does

• Loads and processes Berkeley Earth fixed-width temperature anomaly data for Canada (1960–2024)
• Converts monthly anomalies to annual absolute temperatures using a -4.86°C baseline (1951–1980 reference period)
• Applies three trend methods: 11-year centred running mean, decadal averages, and linear regression
• Projects Canada's average temperature to 2050 (linear regression R² = 0.89)
• Analyses radiative forcing across albedo values 0.20–0.40 using the Stefan-Boltzmann equation
• Merges temperature data with NOAA global CO₂ concentrations and plots the relationship

Key results

Metric	Value
Warming since 1960	~2.8°C
Projected temperature 2050 (linear)	~24.6°C
Radiative forcing range (1960–2024)	3.2–3.8 W/m²
Regression R²	0.89

---

2. Dynamic Carbon Cycle Model

Notebook: carbon_cycle_model.ipynb

What it does

• Implements piecewise-linear emission functions for land-use change (from 1750) and fossil fuels (from 1850)
• Builds a six-pool carbon model: Atmosphere, Biosphere, Soils, Surface Ocean, Deep Ocean, Geosphere
• Solves the ODE system using scipy.integrate.odeint (Runge-Kutta)
• Runs three emission scenarios to 2100:
  • Business-as-Usual: ~700+ ppm CO₂
  • Moderate Mitigation (50% cut by 2050): ~560 ppm
  • Aggressive Mitigation (80% cut by 2050): ~480 ppm
• Sensitivity analysis: ocean uptake rate ±15%, land photosynthesis ±10%

Model parameters

Parameter	Value	Description
β	0.4	CO₂ fertilisation factor
f₀	62 PgC yr⁻¹	Pre-industrial biospheric uptake
k_ψ	0.07 yr⁻¹	Ocean outgassing rate constant

Key results

• Ocean absorbs ~30% of cumulative emissions; land sink weakens over time
• Returning to 300 ppm by 2100 requires net-negative emissions
• Ocean uptake rate is the dominant model uncertainty (±15% on projected CO₂)

---

Requirements

numpy pandas matplotlib scipy sklearn jupyter pip install numpy pandas matplotlib scipy scikit-learn jupyter

Data sources

• Berkeley Earth temperature data: https://berkeleyearth.org/high-resolution-data-access-page/
• NOAA GML CO₂ concentrations: https://gml.noaa.gov/ccgg/trends/gl_data.html

---

How to run

1. Place canada-TAVG-monthly.txt and co2_annmean_gl.txt in the same directory as the notebooks
2. jupyter notebook
3. Run canada_climate_analysis.ipynb then carbon_cycle_model.ipynb