back
loading skill details...
SAR 森林生物量估算:由后向散射 σ⁰ 用线性/饱和经验模型反演森林地上生物量 AGB (t/ha),支持 C/L 波段与地面样本标定
--- name: geoskill-sar-forest-biomass description: 'SAR 森林生物量估算:由后向散射 σ⁰ 用线性/饱和经验模型反演森林地上生物量 AGB (t/ha),支持 C/L 波段与地面样本标定' --- # SAR森林生物量估算 | SAR Forest Biomass Estimation Estimates forest above-ground biomass (AGB, t/ha) from the SAR backscatter coefficient σ⁰. Two empirical relationships are implemented: - **Linear model**: σ⁰_dB = m·AGB + c, suitable for low-biomass, unsaturated ranges. - **Saturation model**: AGB = AGB_sat·(1 − e^(−k·σ⁰_lin)), capturing the physical behavior of SAR backscatter saturating as biomass increases (sensitivity decreases in high-AGB areas). `--calibration` accepts a ground-sample CSV (columns `sigma0,agb`) for coefficient fitting: least squares for the linear model and nonlinear least squares (scipy curve_fit) for the saturation model. Without calibration, built-in default coefficients per band (C / L) are used; the L band has stronger penetration and a higher saturation biomass. ## Application Scenarios / 应用场景 - Forest carbon stock estimation and carbon sink monitoring - Forest resource surveys and biomass mapping - REDD+ and climate change research ## Dependencies / 依赖 ```bash pip install 'numpy' 'rasterio' 'scipy' 'scikit-learn' ``` ## Usage / 使用方法 ### Example 1 (synthetic data, offline, L-band saturation model) ```bash python geoskill-sar-forest-biomass.py --bbox 110.0 22.0 111.0 23.0 --band l --synthetic --output-dir ./out ``` ### Example 2 (C-band linear model) ```bash python geoskill-sar-forest-biomass.py --bbox 110.0 22.0 111.0 23.0 --band c --model linear --synthetic --output-dir ./out ``` ### Example 3 (real σ⁰ imagery + ground-sample calibration) ```bash python geoskill-sar-forest-biomass.py --input sigma0_db.tif --band c --model linear --calibration samples.csv --output-dir ./out ``` ### Example 4 (saturation model calibration) ```bash python geoskill-sar-forest-biomass.py --input sigma0_db.tif --band l --model saturation --calibration samples.csv --output-dir ./out ``` ## Output / 输出 | File | Format | Description | |---|---|---| | `forest_biomass.tif` | GeoTIFF | Retrieved AGB (t/ha) | | `biomass_report.json` | JSON | Model coefficients, statistics, synthetic validation (RMSE/correlation) | | `output-manifest.json` | JSON | Run manifest | ## Data Source / 数据源 / Source Local σ⁰ (dB) GeoTIFF plus optional calibration CSV, or a simulated scene generated by forward modeling of an AGB field via `--synthetic`. ## Privacy / 隐私声明 / Privacy - Runs offline by default; `--synthetic` mode is fully network-free. - All processing is done locally; no user data is uploaded. ## License / License MIT --- <!-- ===== 中文原文 (Chinese Original) ===== --> --- name: geoskill-sar-forest-biomass description: 'SAR 森林生物量估算:由后向散射 σ⁰ 用线性/饱和经验模型反演森林地上生物量 AGB (t/ha),支持 C/L 波段与地面样本标定' --- # SAR森林生物量估算 | SAR Forest Biomass Estimation 从 SAR 后向散射系数 σ⁰ 估算森林地上生物量 AGB(above-ground biomass, t/ha)。 实现两类经验关系: - **线性模型**:σ⁰_dB = m·AGB + c,适用于低生物量、未饱和区间。 - **饱和模型**:AGB = AGB_sat·(1 − e^(−k·σ⁰_lin)),刻画 SAR 后向散射随生物量 增加趋于饱和的物理特征(高 AGB 区敏感度下降)。 支持 `--calibration` 提供地面样本 CSV(列 `sigma0,agb`)做系数拟合:线性用最小 二乘,饱和用非线性最小二乘(scipy curve_fit)。无标定时使用按波段(C / L)内置的 默认系数,L 波段穿透性强、饱和生物量更高。 ## 应用场景 - 森林碳储量估算、碳汇监测 - 森林资源调查与生物量制图 - REDD+ 与气候变化研究 ## 依赖 ```bash pip install 'numpy' 'rasterio' 'scipy' 'scikit-learn' ``` ## 使用方法 ### 示例 1(合成数据,离线,L 波段饱和模型) ```bash python geoskill-sar-forest-biomass.py --bbox 110.0 22.0 111.0 23.0 --band l --synthetic --output-dir ./out ``` ### 示例 2(C 波段线性模型) ```bash python geoskill-sar-forest-biomass.py --bbox 110.0 22.0 111.0 23.0 --band c --model linear --synthetic --output-dir ./out ``` ### 示例 3(真实 σ⁰ 影像 + 地面样本标定) ```bash python geoskill-sar-forest-biomass.py --input sigma0_db.tif --band c --model linear --calibration samples.csv --output-dir ./out ``` ### 示例 4(饱和模型标定) ```bash python geoskill-sar-forest-biomass.py --input sigma0_db.tif --band l --model saturation --calibration samples.csv --output-dir ./out ``` ## 输出 | 文件 | 格式 | 说明 | |---|---|---| | `forest_biomass.tif` | GeoTIFF | 反演 AGB (t/ha) | | `biomass_report.json` | JSON | 模型系数、统计、合成验证 (RMSE/相关) | | `output-manifest.json` | JSON | 运行清单 | ## 数据源 / Source 本地 σ⁰ (dB) GeoTIFF + 可选标定 CSV,或 `--synthetic` AGB 场正演的模拟场景。 ## 隐私声明 / Privacy - 默认离线运行,`--synthetic` 模式完全无网络。 - 所有处理在本地完成,不上传用户数据。 ## License MIT
don't have the plugin yet? install it then click "run inline in claude" again.