🌍 CARBONICA

Carbon Accounting and Regulatory Budget Observatory for Networked Integrated Carbon Assessment

"Weighing the Breath of the Earth." — Samir Baladi, March 2026

---

Overview

CARBONICA is a physically rigorous eight-parameter Earth system science framework for real-time quantification of global carbon cycle dynamics, natural sink capacity, and the critical threshold at which Earth's self-regulating biogeochemical systems approach irreversible saturation.

The framework integrates all governing parameters — from photosynthetic quantum efficiency at the leaf scale to permafrost thaw flux at the continental scale — into a single composite metric: the Planetary Carbon Saturation Index (PCSI).

Current PCSI (2025): 0.78 / 1.00 — Transitional stress zone, accelerating toward critical threshold.

The PCSI has risen from 0.31 (1960) to 0.78 (2025) at an accelerating rate of 0.012 units/year — three times the 1960–1990 rate — driven by simultaneous intensification of three active positive feedback loops.

---

Table of Contents

• Scientific Context
• The Eight Parameters
• The Planetary Carbon Saturation Index (PCSI)
• Installation
• Quick Start
• Project Structure
• Data Sources
• Key Results
• carbonica_engine.py Modules
• Reproducing the Analysis
• Citation
• Author
• License

---

Scientific Context

Carbon is the molecular backbone of life and the primary radiative forcing agent of Earth's climate. The global carbon cycle has never in its geological history experienced a perturbation at the current rate: atmospheric CO₂ is rising at 2.4 ppm/year — approximately 24,000× faster than during typical interglacial periods.

Every natural carbon sink on Earth is showing signs of stress, saturation, or reversal:

• The Amazon rainforest has entered net carbon source status in its southeastern sector.
• Arctic permafrost is thawing at rates exceeding model projections.
• The Revelle Factor has increased from 9.1 (pre-industrial) to 12.4 (2025), indicating a 36% reduction in ocean buffer capacity.
• Photosynthetic quantum yield Φ_q is declining at −0.9%/decade globally.

CARBONICA is designed to track all eight of these parallel, interacting stress signals in an integrated framework validated against a 65-year observational baseline (1960–2025).

---

The Eight Parameters

#	Parameter	Symbol	Role	Critical Threshold
1	Net Primary Productivity	NPP	Terrestrial photosynthetic carbon uptake	< 52.0 PgC/yr
2	Oceanic Carbon Sink Strength	S_ocean	Air-sea CO₂ exchange	< −1.5 PgC/yr (weakening alarm)
3	Atmospheric CO₂ Growth Rate	G_atm	Net source-sink imbalance	≥ 3.5 ppm/yr
4	Permafrost Thaw Flux	F_perma	Frozen carbon reserve release	≥ 2.8 PgC/yr (self-sustaining)
5	Carbon Buffer Capacity	β (= 1/R)	Ocean carbonate buffer chemistry	≤ 1/14.0 = 0.071
6	Soil Carbon Residence Time	τ_soil	Stability of terrestrial carbon reservoir	< 18 yr (net source)
7	Anthropogenic Emission Factor	E_anth	Direct human perturbation term	Net-zero by ~2050
8	Photosynthetic Quantum Yield	Φ_q	Biophysical solar-to-carbon efficiency	< 0.040 (severe stress)

---

The Planetary Carbon Saturation Index (PCSI)

PCSI = w₁·NPP* + w₂·S*_ocean + w₃·G*_atm + w₄·F*_perma + w₅·β* + w₆·τ*_soil + w₇·E*_anth + w₈·Φ*_q

Each parameter is normalized to [0, 1] where 0 = pre-industrial baseline and 1 = defined critical saturation threshold.

Optimized weights (PCA-regularized regression, 1960–2000 training period):

w₁ NPP	w₂ S_ocean	w₃ G_atm	w₄ F_perma	w₅ β	w₆ τ_soil	w₇ E_anth	w₈ Φ_q
0.16	0.18	0.20	0.19	0.12	0.07	0.05	0.03

PCSI Interpretation:

Range	Status
PCSI < 0.55	🟢 Stable — sinks dominating
0.55 – 0.80	🟡 Transitional stress — sink weakening detectable
> 0.80	🔴 Critical — self-reinforcing feedbacks emerging
= 1.00	☠️ Defined saturation — irreversible runaway carbon feedback

Validation against the 65-year Keeling Curve: r² = 0.947 (prospective 2001–2025 test period).

---

Installation

Python Package

pip install carbonica

From Source

git clone https://gitlab.com/gitdeeper9/carbonica.git
cd carbonica
pip install -e .

Requirements

python >= 3.9
numpy >= 1.24
scipy >= 1.10
pandas >= 2.0
netCDF4 >= 1.6
h5py >= 3.8
matplotlib >= 3.7
xarray >= 2023.1

---

Quick Start

from carbonica import CarbonBudget, OceanSinkModel, PermafrostEngine, QuantumYieldTracker
from carbonica.pcsi import PCSI

# Initialize the full eight-parameter system
budget = CarbonBudget(baseline_year=1960, end_year=2025)
budget.load_observations()   # auto-fetches NOAA, SOCAT, GCP, GTN-P streams

# Compute current PCSI
index = PCSI(budget)
print(f"Current PCSI (2025): {index.current:.3f}")
# → Current PCSI (2025): 0.780

# Run SSP3-7.0 projection to 2070
projection = index.project(scenario="SSP3-7.0", end_year=2070, n_ensemble=10000)
print(f"PCSI critical threshold crossing: {projection.threshold_year(0.90)}")
# → PCSI critical threshold crossing: 2047–2053

# Inspect individual parameters
print(budget.permafrost.F_perma_2025)     # → 1.71 ± 0.40 PgC/yr
print(budget.ocean.revelle_factor_2025)   # → 12.4
print(budget.biosphere.phi_q_trend)       # → -0.009 /decade

---

Project Structure

carbonica/
│
├── README.md                          # This file
├── LICENSE
├── pyproject.toml
├── setup.cfg
│
├── carbonica/                         # Core Python package
│   ├── __init__.py
│   ├── carbonica_engine.py            # Main engine — all four core modules
│   ├── pcsi.py                        # Planetary Carbon Saturation Index
│   │
│   ├── modules/
│   │   ├── carbon_budget.py           # CarbonBudget: master dC_atm/dt integrator
│   │   ├── ocean_sink.py              # OceanSinkModel: Wanninkhof (2014) gas transfer
│   │   ├── permafrost_engine.py       # PermafrostEngine: Q₁₀ decomposition model
│   │   └── quantum_yield_tracker.py   # QuantumYieldTracker: GOSAT/OCO-2 SIF retrieval
│   │
│   ├── data/
│   │   ├── loaders/
│   │   │   ├── noaa_loader.py         # NOAA Mauna Loa CO₂ record
│   │   │   ├── socat_loader.py        # SOCAT surface ocean pCO₂ atlas
│   │   │   ├── gcp_loader.py          # Global Carbon Project annual budgets
│   │   │   ├── gtnp_loader.py         # GTN-P permafrost borehole network
│   │   │   ├── glodap_loader.py       # GLODAP ocean carbonate chemistry
│   │   │   ├── modis_loader.py        # MODIS MOD17 NPP + fAPAR
│   │   │   └── sif_loader.py          # GOSAT + OCO-2 solar-induced fluorescence
│   │   │
│   │   └── cache/                     # Local NetCDF4 / HDF5 observation cache
│   │       ├── keeling_1960_2025.nc
│   │       ├── socat_v2023.nc
│   │       ├── gtnp_boreholes.nc
│   │       └── glodap_v2.nc
│   │
│   ├── models/
│   │   ├── box_model.py               # 5-reservoir ODE system (Appendix B)
│   │   ├── lue_model.py               # Light Use Efficiency NPP model
│   │   ├── revelle.py                 # Revelle Factor time-series derivation
│   │   └── ssp_scenarios.py           # SSP1-1.9 / SSP3-7.0 / SSP5-8.5 forcing
│   │
│   ├── statistics/
│   │   ├── monte_carlo.py             # 10,000-member PCSI uncertainty ensemble
│   │   ├── sem.py                     # Structural Equation Modelling (DAG)
│   │   ├── pca_regression.py          # PCA-regularized weight optimization
│   │   └── cusum.py                   # CUSUM change-point detection
│   │
│   └── visualization/
│       ├── pcsi_dashboard.py          # Real-time web dashboard backend
│       ├── parameter_plots.py         # Eight-parameter time series plots
│       ├── correlation_matrix.py      # Heatmap renderer
│       └── projection_plot.py         # SSP scenario fan charts
│
├── notebooks/                         # Jupyter notebooks — all manuscript figures
│   ├── 01_keeling_validation.ipynb    # Fig 1: PCSI vs Keeling Curve 1960–2025
│   ├── 02_revelle_factor.ipynb        # Fig 2: Revelle Factor time series
│   ├── 03_permafrost_flux.ipynb       # Fig 3: F_perma reconstruction & projection
│   ├── 04_quantum_yield.ipynb         # Fig 4: Global Φ_q trend (GOSAT + OCO-2)
│   ├── 05_correlation_matrix.ipynb    # Fig 5: Eight-parameter correlation heatmap
│   ├── 06_amazon_case_study.ipynb     # Fig 6: Amazon carbon pivot 2018–2023
│   ├── 07_arctic_subsystem.ipynb      # Fig 7: Arctic permafrost spatial analysis
│   ├── 08_ssp_projections.ipynb       # Fig 8: PCSI SSP scenario fan (2025–2070)
│   ├── 09_sem_pathways.ipynb          # Fig 9: SEM carbon feedback DAG
│   ├── 10_monte_carlo_uncertainty.ipynb
│   ├── 11_tipping_points.ipynb
│   └── 12_policy_budget_correction.ipynb
│
├── data/                              # Raw & processed archival data
│   ├── keeling/                       # NOAA Mauna Loa monthly CO₂ (1958–2025)
│   ├── socat/                         # SOCAT v2023 pCO₂ cruise tracks
│   ├── gcp/                           # Global Carbon Project 2023 budget tables
│   ├── gtnp/                          # GTN-P borehole temperatures (1,200+ sites)
│   ├── glodap/                        # GLODAP v2 ocean DIC / alkalinity
│   ├── modis/                         # MODIS MOD17 NPP 500 m / 8-day mosaics
│   ├── gosat/                         # GOSAT SIF L2 retrievals (2009–2025)
│   └── oco2/                          # OCO-2 SIF retrievals (2014–2025)
│
├── results/                           # Model outputs (gitignored large files)
│   ├── pcsi_timeseries_1960_2025.nc   # Full PCSI parameter inversion archive
│   ├── pcsi_ensemble_ssp370.h5        # 10,000-member Monte Carlo ensemble
│   ├── pcsi_ensemble_ssp119.h5
│   └── pcsi_ensemble_ssp585.h5
│
├── docs/                              # Documentation
│   ├── api/                           # Auto-generated API reference (Sphinx)
│   ├── theory/
│   │   ├── box_model.md               # ODE system derivation (Appendix B)
│   │   ├── revelle_factor.md          # Revelle Factor chemistry
│   │   └── pcsi_formulation.md        # PCSI weight optimization methodology
│   └── changelog/
│       └── CHANGELOG.md
│
└── tests/                             # Unit & integration tests
    ├── test_carbon_budget.py
    ├── test_ocean_sink.py
    ├── test_permafrost_engine.py
    ├── test_quantum_yield.py
    ├── test_pcsi.py
    └── test_box_model.py

---

Data Sources

Parameter	Primary Source	Record Length	Resolution
G_atm (CO₂ growth rate)	NOAA Mauna Loa Observatory	1960–2025 (65 yr)	Daily; ±0.05 ppm
S_ocean (air-sea flux)	SOCAT Surface Ocean CO₂ Atlas v2023	1970–2025 (55 yr)	1°×1° monthly
NPP (productivity)	MODIS MOD17 + NASA OCO-2 SIF	2000–2025 (MODIS)	500 m / 8-day
F_perma (permafrost)	GTN-P Global Terrestrial Network	1990–2025 (35 yr)	1,200+ boreholes
β (Revelle Factor)	GLODAP Global Ocean Data Analysis Project v2	1972–2025	Full-depth carbonate
τ_soil (residence time)	ISCN + FLUXNET (900+ eddy covariance sites)	1980–2025	71,000+ profiles
E_anth (anthropogenic)	Global Carbon Project + IEA CO₂ Statistics	1960–2025 (65 yr)	Annual, country-level
Φ_q (quantum yield)	GOSAT SIF (2009–2025) + OCO-2 SIF (2014–2025)	16 yr	2°×2° monthly

---

Key Results

PCSI Decadal Evolution

Decade	PCSI	G_atm (ppm/yr)	S_ocean (PgC/yr)	F_perma (PgC/yr)
1960–1970	0.31	0.90	−1.21	~0.0
1971–1980	0.38	1.28	−1.44	~0.0
1981–1990	0.43	1.51	−1.78	~0.01
1991–2000	0.49	1.63	−1.93	0.05
2001–2010	0.58	1.88	−2.24	0.31
2011–2020	0.68	2.19	−2.71	0.98
2021–2025	0.78	2.38	−3.08	1.71

Critical Findings

• Permafrost feedback: F_perma has risen from near-zero (1990) to 1.71 ± 0.40 PgC/yr (2025) — constituting 4.3% of global emissions. Abrupt thaw (thermokarst) contributes 31% of current F_perma despite covering only 5% of permafrost area.
• Revelle Factor acceleration: Ocean buffer capacity declining at 0.067 R-units/year (2.5× the 1960–1990 rate). Projected to reach 14.0 by 2050, reducing ocean uptake efficiency from 28% to ~18%.
• Quantum yield decline: Global Φ_q declining at −0.9%/decade (2009–2025), with tropical ecosystems at −2.1%/decade.
• Policy implication: Achieving net-zero human emissions by 2050 would not prevent PCSI exceeding 0.85 by 2055, because autonomous permafrost and soil carbon feedbacks will contribute 2.5–4.0 PgC/yr through 2060. The effective remaining carbon budget for 1.5°C is 15–25 PgC smaller than the IPCC AR6 estimate.

PCSI Projection

Scenario	PCSI = 0.90 Threshold Crossing
SSP1-1.9 (aggressive mitigation)	~2055–2067
SSP3-7.0 (current trajectory)	2047–2053
SSP5-8.5 (high emissions)	~2041–2046

---

carbonica_engine.py Modules

CarbonBudget

Integrates the master dC_atm/dt equation in real time from NOAA, SOCAT, GCP, and GTN-P data streams. Outputs parameter time series at monthly resolution; propagates uncertainty through Monte Carlo sampling (10,000 ensemble members). Detects rate-of-change anomalies using CUSUM change-point detection.

OceanSinkModel

Implements the Wanninkhof (2014) gas transfer parameterization globally on a 1°×1° grid using ERA5 wind fields and SOCAT pCO₂ data. Tracks Revelle Factor evolution across 12 ocean basin sub-regions. Real-time β monitoring with 30-day forecast of ocean uptake efficiency.

PermafrostEngine

Couples GTN-P active layer depth to the Q₁₀ decomposition model. Separates gradual thaw from abrupt thaw (thermokarst) signals using GRACE-FO lake area change. Projects F_perma under SSP scenarios with explicit treatment of tipping point dynamics (bistable permafrost state).

QuantumYieldTracker

Derives global Φ_q maps from GOSAT and OCO-2 SIF retrievals using the Frankenberg et al. (2011) fluorescence-photosynthesis algorithm. Detects drought and heat stress episodes as Φ_q anomalies below the biome-specific climatological mean. Feeds directly into the NPP forward model.

---

Reproducing the Analysis

All manuscript figures are fully reproducible from the provided notebooks:

# Clone repository
git clone https://gitlab.com/gitdeeper9/carbonica.git
cd carbonica

# Install dependencies
pip install -e ".[notebooks]"

# Launch Jupyter
jupyter lab notebooks/

# Or run all notebooks headlessly
jupyter nbconvert --to notebook --execute notebooks/*.ipynb

All 12 notebooks reproduce the corresponding manuscript figures and statistical outputs without any external dependencies beyond the data cached in data/.

---

Citation

If you use CARBONICA in your research, please cite:

@article{baladi2026carbonica,
  title     = {{CARBONICA}: Advanced Planetary Carbon Accounting \& Feedback Dynamics —
               A Multi-Parameter Earth System Science Framework for Real-Time
               Quantification of Global Carbon Cycle Dynamics, Sink Saturation,
               and Planetary Self-Regulation Thresholds},
  author    = {Baladi, Samir},
  journal   = {Nature Climate Change},
  year      = {2026},
  month     = {March},
  doi       = {10.5281/zenodo.18995446},
  url       = {https://doi.org/10.5281/zenodo.18995446}
}

---

Author

Samir Baladi Interdisciplinary AI Researcher Ronin Institute / Rite of Renaissance

• 📧 gitdeeper@gmail.com
• 🔬 ORCID: 0009-0003-8903-0029
• 🌐 Dashboard: carbonica.netlify.app
• 💻 GitHub: github.com/gitdeeper9/carbonica
• 🦊 GitLab: gitlab.com/gitdeeper9/carbonica

---

License

This project is fully open-access. Code, datasets, PCSI projection ensembles (NetCDF4/HDF5), and all supplementary materials are archived at:

• Zenodo: https://doi.org/10.5281/zenodo.18995446
• PyPI: pip install carbonica
• Dashboard: https://carbonica.netlify.app

---

CARBONICA v1.0.0 · Submitted to Nature Climate Change, March 2026