geoskill-precision-forestry-monitoring

CHM tree height and crown width allometric volume, NDVI/NDRE health grading and SAR biomass fusion for precision forestry

精准林业遥感监测 | Precision Forestry Monitoring

Integrates multi-source remote sensing for stand-scale precision forestry monitoring, covering four levels — individual tree, stand, health and biomass — to support forest resource inventory, carbon sink estimation and management decisions.

Tree height / individual trees: CHM = DSM − DTM, with local peak detection (non-maximum suppression) extracting individual tree locations and heights. Crown width is used to invert diameter at breast height (DBH = k·CW), and the allometric equation V = a·DBH^b·H^c estimates individual-tree volume, which is summed into stand volume. Canopy closure is the proportion of canopy pixels whose CHM exceeds a threshold (∈[0,1]). Health is graded by a combination of NDVI and NDRE (healthy / moderate / stressed). Biomass is estimated from SAR backscatter σ⁰ (dB) → linear → power-function empirical relationships, and management recommendations are generated from canopy closure / stand volume / health.

Dependencies / 依赖

bash

pip install 'numpy' 'rasterio' 'scipy' 'scikit-learn'

Usage / 使用方法

Basic usage

bash

python geoskill-precision-forestry-monitoring.py --bbox 116.0 39.0 117.0 40.0 [other options]

Example 1 (synthetic stand, offline)

bash

python geoskill-precision-forestry-monitoring.py --bbox 116 39 117 40 --synthetic --output-dir ./out

Example 2 (real 6-band imagery (DSM/DTM/Red/NIR/RedEdge/SAR))

bash

python geoskill-precision-forestry-monitoring.py --input forest.tif --output-dir ./out

Example 3 (health grading using NDVI only)

bash

python geoskill-precision-forestry-monitoring.py --input forest.tif --health-method ndvi --output-dir ./out

Example 4 (increase the minimum detection height and window)

bash

python geoskill-precision-forestry-monitoring.py --input forest.tif --min-height 5 --footprint 7 --output-dir ./out

Example 5 (adjust the canopy closure threshold)

bash

python geoskill-precision-forestry-monitoring.py --input forest.tif --closure-threshold 3 --output-dir ./out

Output / 输出

File	Format	Description
chm.tif	GeoTIFF	Canopy height model (tree height, m)
health_grade.tif	GeoTIFF	Health grade (3 healthy / 2 moderate / 1 stressed / 0 bare ground)
biomass_t_ha.tif	GeoTIFF	SAR-derived above-ground biomass (t/ha)
canopy_mask.tif	GeoTIFF	Canopy cover mask
trees.geojson	GeoJSON	Individual tree locations (height / crown width / DBH / volume)
forestry_report.json	JSON	Stand statistics + management recommendations
output-manifest.json	JSON	Run manifest

Data Source / 数据源 / Source

Multi-band GeoTIFF with band order DSM / DTM / Red / NIR / RedEdge / SAR. Alternatively, use --synthetic to generate physically consistent simulated data (fully offline).

Privacy / 隐私声明 / Privacy

• Runs offline by default; --synthetic mode requires no network at all.
• All processing is performed locally; user data is never uploaded.

License / License

MIT

---

---

name: geoskill-precision-forestry-monitoring description: 'CHM tree height and crown width allometric volume, NDVI/NDRE health grading and SAR biomass fusion for precision forestry'

精准林业遥感监测 | Precision Forestry Monitoring

融合多源遥感实现林分尺度精准林业监测，覆盖“单木—林分—健康—生物量”四个层次，支撑森林资源调查、碳汇估算与经营决策。

树高/单木：CHM = DSM − DTM，局部峰值检测（非极大值抑制）提取单木位置与树高；由冠幅反演胸径 (DBH = k·CW)，再用异速生长方程 V = a·DBH^b·H^c 估单木蓄积量并累加为林分蓄积量。郁闭度取 CHM 高于阈值的冠层像元占比（∈[0,1]）。健康由 NDVI 与 NDRE 组合分级（健康/中等/胁迫）。生物量由 SAR 后向散射 σ⁰(dB)→线性→幂函数经验关系估算，并据郁闭度/蓄积量/健康生成经营建议。

依赖

bash

pip install 'numpy' 'rasterio' 'scipy' 'scikit-learn'

使用方法

基本用法

bash

python geoskill-precision-forestry-monitoring.py --bbox 116.0 39.0 117.0 40.0 [其他参数]

示例 1（合成林分，离线）

bash

python geoskill-precision-forestry-monitoring.py --bbox 116 39 117 40 --synthetic --output-dir ./out

示例 2（真实 6 波段影像（DSM/DTM/Red/NIR/RedEdge/SAR））

bash

python geoskill-precision-forestry-monitoring.py --input forest.tif --output-dir ./out

示例 3（仅用 NDVI 分级健康）

bash

python geoskill-precision-forestry-monitoring.py --input forest.tif --health-method ndvi --output-dir ./out

示例 4（调高最小检测树高与窗口）

bash

python geoskill-precision-forestry-monitoring.py --input forest.tif --min-height 5 --footprint 7 --output-dir ./out

示例 5（调整郁闭度阈值）

bash

python geoskill-precision-forestry-monitoring.py --input forest.tif --closure-threshold 3 --output-dir ./out

输出

文件	格式	说明
chm.tif	GeoTIFF	冠层高度模型（树高，m）
health_grade.tif	GeoTIFF	健康分级（3 健康 / 2 中等 / 1 胁迫 / 0 裸地）
biomass_t_ha.tif	GeoTIFF	SAR 反演地上生物量（t/ha）
canopy_mask.tif	GeoTIFF	冠层覆盖掩膜
trees.geojson	GeoJSON	单木位置（树高/冠幅/胸径/蓄积量）
forestry_report.json	JSON	林分统计 + 经营建议
output-manifest.json	JSON	运行清单

数据源 / Source

多波段 GeoTIFF，波段顺序 DSM / DTM / Red / NIR / RedEdge / SAR。 或使用 --synthetic 生成物理一致的模拟数据（完全离线）。

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