| name | ecosystem-services-assessment |
| description | Maps and quantifies ecosystem services including carbon stocks, water yield, soil erosion, and habitat quality with trade-off analysis. Use this skill when the user mentions ecosystem services, InVEST models, ES mapping, carbon sequestration, water yield estimation, RUSLE erosion modeling, habitat quality, pollination services, trade-off analysis, PES (payments for ecosystem services), natural capital, or ES valuation. |
| skill_version | 1.0.0 |
Skill: ecosystem-services-assessment
Domain: ES indicators · Provisioning · Regulating · Cultural · Trade-offs
Phase: 3 — Specialist
Used by: assess-ecosystem-services
Purpose
Guides the agent through the quantification and spatial representation of ecosystem services (ES): selecting appropriate indicators, computing biophysical ES estimates, mapping ES supply, and analysing trade-offs and synergies between services.
When to Invoke
- Quantifying ecosystem services across a landscape or watershed
- Mapping ES supply and demand
- Analysing trade-offs between conservation and production services
- Supporting payments for ecosystem services (PES) design
- Environmental impact assessments requiring ES valuation
Inputs
| Input | Format | Required |
|---|
| Land cover / land use map | GeoTIFF, SHP | Yes |
| Biophysical data (rainfall, DEM, soil, biomass) | GeoTIFF, CSV | Yes |
| Socioeconomic data (population, demand) | GeoTIFF, CSV | Optional |
| Study area polygon | SHP, GPKG | Yes |
Outputs
| Output | Description |
|---|
es_indicator_maps/ | One raster per ES indicator |
es_summary_table.csv | ES value per land cover class |
tradeoff_matrix.csv | Pairwise ES correlation matrix |
tradeoff_plot.png | Scatter plots or heatmap of ES trade-offs |
es_report.md | Full ES assessment narrative |
Steps
1. Define the ES Portfolio
Select relevant services for the context:
| Category | Examples |
|---|
| Provisioning | Timber, water, food, fibre, genetic resources |
| Regulating | Carbon sequestration, water regulation, erosion control, pollination |
| Cultural | Recreation, aesthetic value, spiritual significance |
| Supporting | Habitat, nutrient cycling (underlying processes) |
2. Select Indicators and Methods per Service
- Carbon: Above-ground biomass from REDD+ datasets or allometric models; soil carbon from SoilGrids
- Water regulation: Curve number (CN) approach; InVEST Seasonal Water Yield
- Erosion control: Revised Universal Soil Loss Equation (RUSLE); C-factor from land cover
- Pollination: Distance-weighted bee habitat index from land cover
- Recreation: Proximity index to natural areas weighted by accessibility
3. Compute Biophysical ES Values
- Apply selected method per service; store as raster layer
- Validate against field measurements or published benchmarks where possible
- Document all input parameters and data sources
4. Aggregate by Land Cover Class
- Zonal statistics: mean/total ES value per land cover polygon or raster class
- Build summary table: rows = land cover classes, columns = ES indicators
- Normalise to 0–1 for cross-service comparison
5. Trade-off and Synergy Analysis
- Compute pairwise Spearman correlations across pixels or land cover units
- Positive correlations = synergies; negative = trade-offs
- Visualise as correlation heatmap and scatter plots
- Identify land cover classes that maximise multiple services simultaneously
6. Beneficiary Mapping (optional)
- Map ES demand using population density, agricultural areas, or water intake points
- Overlay supply and demand to identify supply-demand gaps
Decision Points
| Condition | Diagnosis | Recommended Action |
|---|
| Land cover map accuracy < 85% | Classification error propagates into ES estimates | Perform uncertainty analysis; include classification error as confidence range in ES outputs |
| Land cover data > 5 years old | Land use change not captured — ES estimates may be outdated | Flag temporal gap; recommend updated land cover if available; note limitation in report |
| InVEST NoData fraction > 30% in study area | CRS or extent mismatch, or missing input raster | Check CRS alignment and extent; rerun after fixing spatial inputs |
| ES trade-off between two services is strongly negative (r < -0.7) | Managing for one service degrades the other | Report trade-off explicitly; do not recommend single-service optimisation without acknowledging cost |
Key Decisions to Document
- ES portfolio selection rationale
- Method and data source per service
- Normalisation method for cross-service comparison
- Trade-off analysis unit (pixel, land cover class, watershed)
Tools and Libraries
R: terra, dplyr, ggplot2, corrplot
Python: rasterio, geopandas, seaborn
Dedicated: InVEST (Stanford Natural Capital Project), ARIES, Co$ting Nature
Resources
resources/es-indicator-reference.md — indicator definitions per serviceresources/invest-parameter-guide.md — InVEST model configurationresources/rusle-coefficients.md — RUSLE factor lookup tablesexamples/ — worked carbon + water regulation example
Notes
- ES assessment does not require monetary valuation; biophysical indicators are often sufficient and more defensible
- InVEST is the most widely used open-source platform; use it unless a specific method is required
- Always report uncertainty in ES estimates, especially for carbon stocks
