Geoskill: Forest Carbon Estimate

forest-carbon-estimate

Estimate forest carbon stock from remote sensing data using multiple methods with uncertainty analysis.

Features

• BEF Method: Biomass Expansion Factor from AGB
• Allometric Equations: AGB = a × H^b from forest height
• IPCC Tier 1/2: Default factors by forest type
• Monte Carlo Uncertainty: Propagate input uncertainties
• Raster Processing: Direct GeoTIFF input/output
• Tabular Processing: CSV with plot-level data
• Multiple Forest Types: Tropical, temperate, boreal, mangrove

Usage

# Single-point estimate (allometric)
python scripts\forest-carbon-estimate.py estimate --method allometric --height 15 --forest-type tropical

# Single-point estimate (BEF)
python scripts\forest-carbon-estimate.py estimate --method bef --agb 200 --forest-type temperate

# IPCC Tier 1 default
python scripts\forest-carbon-estimate.py estimate --method ipcc --forest-type boreal --area-ha 100

# Raster processing
python scripts\forest-carbon-estimate.py estimate --input height.tif --method allometric --output carbon.tif

# Uncertainty analysis
python scripts\forest-carbon-estimate.py uncertainty --method allometric --height 15 --iterations 5000

# Report from CSV
python scripts\forest-carbon-estimate.py report --input carbon_stock.csv

Parameters

Calculation Chain

AGB (Above-ground biomass)
  ↓
BGB = AGB × root_shoot_ratio (default 0.26)
  ↓
Total biomass = AGB + BGB
  ↓
Carbon stock = Total biomass × carbon_fraction (default 0.47)

Methods

Installation

pip install requests>=2.28.0 tqdm numpy scipy rasterio
# Or: pip install -r scripts/requirements.txt

Dependencies

Data Source

• IPCC Guidelines for National Greenhouse Gas Inventories (2006, 2019 Refinement)
• IPCC EFDB (Emission Factor Database)

BEF Values per Forest Type

Default BEF = 1.32 (temperate mixed forest). Adjust with --bef parameter.

Allometric Coefficients

For AGB = a × H^b (H = forest height in m):

These are DBH-based defaults. For height-based allometry, use --coeff-a and --coeff-b to override.

IPCC Default Wood Density

Default: 0.55 g/cm³. Override with --wood-density.

Method Selection Guidance

Output Units

Carbon stock is reported in Mg C/ha (megagrams of carbon per hectare), equivalent to t C/ha.

For total stock: multiply by area (ha) → total Mg C.

Nodata Handling

Nodata pixels in input rasters are skipped. Output GeoTIFF uses the same nodata value as input. No interpolation is performed on nodata areas.

Custom Parameters

# Custom carbon fraction and root-shoot ratio
python scripts\forest-carbon-estimate.py estimate --method bef --agb 200 --forest-type temperate --carbon-fraction 0.45 --root-shoot-ratio 0.28

Uncertainty Output Structure

{
  "mean": 125.3,
  "std": 18.7,
  "CI95_lower": 88.6,
  "CI95_upper": 162.0
}

Data Acquisition Guidance

Obtain forest height/AGB rasters from:

• Global Forest Watch (https://www.globalforestwatch.org/) — AGB maps
• NASA GEDI (https://gedi.umd.edu/) — Forest height from spaceborne LiDAR
• ESA Biomass Mission — P-band SAR forest height
• National forest inventory — Plot-level AGB data

Validation / Quality Assessment

• Compare with field-measured carbon stock plots
• Cross-validate with IPCC default values for the same forest type
• Check that uncertainty range (CI95) is reasonable (< 50% of mean)
• Report method, forest type, and input data source in publications

Citation

@book{ipcc2006guidelines,
  title={2006 IPCC Guidelines for National Greenhouse Gas Inventories},
  author={{IPCC}},
  year={2006},
  publisher={Institute for Global Environmental Strategies},
  url={https://www.ipcc-nggip.iges.or.jp/public/2006gl/}
}
@book{ipcc2019refinement,
  title={2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories},
  author={{IPCC}},
  year={2019},
  publisher={IPCC},
  url={https://www.ipcc-nggip.iges.or.jp/public/2019rf/}
}

Visualization Guidance

import rasterio
import matplotlib.pyplot as plt
import numpy as np

with rasterio.open("carbon.tif") as src:
    carbon = src.read(1)
    nodata = src.nodata

carbon_plot = np.where(carbon == nodata, np.nan, carbon)

fig, ax = plt.subplots(figsize=(10, 8))
im = ax.imshow(carbon_plot, cmap="Greens", vmin=0, vmax=200)
cbar = plt.colorbar(im, ax=ax, shrink=0.8)
cbar.set_label("Carbon Stock (Mg C/ha)")
ax.set_title("Forest Carbon Stock")
ax.axis("off")
plt.tight_layout()
plt.savefig("carbon_map.png", dpi=200)

Troubleshooting

---

Advanced Usage

Batch Raster Processing

for year in 2020 2021 2022 2023; do
  python scripts\forest-carbon-estimate.py estimate     --input agb_${year}.tif --method allometric     --output carbon_${year}.tif
done

CI/CD Integration (GitHub Actions)

# .github/workflows/carbon-update.yml
name: Forest Carbon Update
on:
  schedule:
    - cron: '0 0 1 1 *'  # Yearly
jobs:
  estimate:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: '3.11'
      - run: pip install numpy rasterio
      - run: |
          python scripts\forest-carbon-estimate.py estimate \
            --input data/agb_latest.tif \
            --method allometric \
            --output data/carbon_latest.tif

PostGIS Raster Import

raster2pgsql -s 4326 -I -C carbon_latest.tif public.forest_carbon | psql -d gis_db

Performance Tips

• --method bef is fastest for regional estimates; --method allometric for species-specific
• Use --carbon-fraction 0.47 to match local species (default 0.47 = IPCC default)
• For large rasters, process in tiles using --window parameter

---

中文说明

基于遥感数据估算森林碳储量，支持 BEF、异速生长方程、IPCC Tier 1/2 三种方法，含蒙特卡洛不确定性分析。

安装

pip install requests>=2.28.0 tqdm numpy scipy rasterio
# 或: pip install -r scripts/requirements.txt

依赖

各森林类型 BEF 值

默认 BEF = 1.32（温带混交林）。使用 --bef 参数调整。

异速生长系数

AGB = a × H^b（H = 树高，单位 m）：

这些是基于 DBH 的默认值。树高异速生长使用 --coeff-a 和 --coeff-b 覆盖。

IPCC 默认木材密度