geoskill-insurance-risk-mapping

Multi-hazard probability times asset value times vulnerability curves to compute expected loss for insurance risk mapping

保险风险制图 | Insurance Risk Mapping

Multi-hazard Expected Annual Loss (EAL) mapping for property / catastrophe insurance, providing a spatial basis for underwriting pricing, loss reserving, and risk diversification.

Core model EAL = Σ_h P_h · Asset · V_h(I_h): P_h = 1/recurrence interval is the annual exceedance probability of hazard h, Asset is the per-pixel asset value, and V_h is the vulnerability curve (linear or Sigmoid) mapping hazard intensity to a loss ratio ∈ [0, 1]. Flood, wind, and seismic losses are computed separately and stacked to yield per-pixel annual expected loss and risk class.

Dependencies / 依赖

bash

pip install 'numpy' 'rasterio' 'scipy'

Usage / 使用方法

Basic usage

bash

python geoskill-insurance-risk-mapping.py --bbox 116.0 39.0 117.0 40.0 [other arguments]

Example 1 (Synthetic scenario, offline)

bash

python geoskill-insurance-risk-mapping.py --bbox 116 39 117 40 --synthetic --output-dir ./out

Example 2 (Real data (Asset/Flood/Wind/Seismic))

bash

python geoskill-insurance-risk-mapping.py --input data.tif --output-dir ./out

Example 3 (Switch to Sigmoid vulnerability curve)

bash

python geoskill-insurance-risk-mapping.py --input data.tif --curve sigmoid --output-dir ./out

Example 4 (Custom risk classes)

bash

python geoskill-insurance-risk-mapping.py --input data.tif --class-breaks 5 50 500 --output-dir ./out

Output / 输出

File	Format	Description
expected_annual_loss.tif	GeoTIFF	Multi-hazard annual expected loss
per_hazard_loss.tif	GeoTIFF	Per-hazard loss (flood / wind / seismic)
risk_class.tif	GeoTIFF	Risk class
risk_report.json	JSON	Total insured value / total loss / loss ratio / class statistics
output-manifest.json	JSON	Run manifest

Data Source / 数据源 / Source

Multi-band GeoTIFF with band order Asset / Flood / Wind / Seismic intensity. Or use --synthetic to generate physically consistent simulated data (fully offline).

Privacy / 隐私声明 / Privacy

• Runs offline by default; --synthetic mode is fully offline with no network access.
• All processing is performed locally; user data is never uploaded.

License / License

MIT

---

---

name: geoskill-insurance-risk-mapping description: 'Multi-hazard probability times asset value times vulnerability curves to compute expected loss for insurance risk mapping'

保险风险制图 | Insurance Risk Mapping

面向财产/巨灾保险的多灾种年期望损失 (Expected Annual Loss, EAL) 制图，为承保定价、准备金与风险分散提供空间依据。

核心模型 EAL = Σ_h P_h · Asset · V_h(I_h)：P_h = 1/重现期为第 h 种灾害年超越概率，Asset 为像元资产价值，V_h 为把灾害强度映射为损失比 ∈[0,1] 的脆弱性曲线（线性或 Sigmoid）。对洪水、风灾、地震分别计算后叠加，得到逐像元年期望损失与风险等级。

依赖

bash

pip install 'numpy' 'rasterio' 'scipy'

使用方法

基本用法

bash

python geoskill-insurance-risk-mapping.py --bbox 116.0 39.0 117.0 40.0 [其他参数]

示例 1（合成场景，离线）

bash

python geoskill-insurance-risk-mapping.py --bbox 116 39 117 40 --synthetic --output-dir ./out

示例 2（真实数据（Asset/Flood/Wind/Seismic））

bash

python geoskill-insurance-risk-mapping.py --input data.tif --output-dir ./out

示例 3（改用 Sigmoid 脆弱性曲线）

bash

python geoskill-insurance-risk-mapping.py --input data.tif --curve sigmoid --output-dir ./out

示例 4（自定义风险分档）

bash

python geoskill-insurance-risk-mapping.py --input data.tif --class-breaks 5 50 500 --output-dir ./out

输出

文件	格式	说明
expected_annual_loss.tif	GeoTIFF	多灾种年期望损失
per_hazard_loss.tif	GeoTIFF	分灾种损失（洪水/风灾/地震）
risk_class.tif	GeoTIFF	风险等级
risk_report.json	JSON	总保额/总损失/损失率/分级统计
output-manifest.json	JSON	运行清单

数据源 / Source

多波段 GeoTIFF，波段顺序 Asset / Flood / Wind / Seismic 强度。 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